|Year : 2019 | Volume
| Issue : 6 | Page : 340-346
Spatial distribution of dentists in Thailand
Wuttikul Thanakanjanaphakdee1, Wongsa Laohasiriwong2, Nattapong Puttanapong3
1 Doctoral of Public Health Program, Faculty of Public Health, Khon Kaen University, Khon Kaen, Thailand
2 Faculty of Public Health, Khon Kaen University, Khon Kaen, Thailand
3 Faculty of Economics, Thammasat University, Bangkok, Thailand
|Date of Web Publication||26-Nov-2019|
Dr. Wongsa Laohasiriwong
Faculty of Public Health, Khon Kaen University, Khon Kaen 40002.
Source of Support: None, Conflict of Interest: None
Background: Thailand has been facing maldistribution of dentists for decades despite its continuous increase in total numbers. Aims: The objective of this study was to analyze spatial inequality of dentist distribution patterns in Thailand. Materials and Methods: Data on dentist in public sector profiles between 2007 and 2015 were retrieved from the Ministry of Public Health, Thailand, to analyze the distribution patterns and spatial relationships. The distribution of dentists was visualized on a map using Quantum Geographic Information System. Global Moran I and local indicators of spatial association were analyzed using GeoDa software to determine the distribution of spatial relationship. Results: The total number of dentists has been increased annually from 2007 to 2015. Initially, the high–high (population-to-dentist ratio) spatial clusters were common in the Northeast (NE) region, whereas there were many low–low spatial clusters in Central region. In 2015, the high–high spatial clusters persisted in the NE region, but the low–low spatial clusters were hardly found in the Central region. Conclusions: Unequal geographical distribution of dentist still existed in Thailand, especially in the NE region. Both supply side concerning administrative factors related to resource allocation, and demand side of socioeconomic development and population changes from migration had influences on inequality in dentist distribution.
Keywords: Dentist, global Moran I, LISA, maldistribution, spatial clustering pattern
|How to cite this article:|
Thanakanjanaphakdee W, Laohasiriwong W, Puttanapong N. Spatial distribution of dentists in Thailand. J Int Oral Health 2019;11:340-6
|How to cite this URL:|
Thanakanjanaphakdee W, Laohasiriwong W, Puttanapong N. Spatial distribution of dentists in Thailand. J Int Oral Health [serial online] 2019 [cited 2020 Jun 4];11:340-6. Available from: http://www.jioh.org/text.asp?2019/11/6/340/271775
| Introduction|| |
Spatial analysis is a contemporary technique used to explore relationships between the environment and health. It has been used in many epidemiological studies, including analysis of the geographic distribution of diseases, association of economic and environmental factors, and finding pathogenic agents.,,,, It can help expand our understanding of disparities in health outcomes of community. Spatial analysis using geographic information system (GIS) residency footprinting is an important analytic tool to ensure health workforce needs as well as to present information in a way that increases its usefulness in dental service planning. Spatial analysis is important for understanding, planning, monitoring, and allocating health resources. However, its application in the field of geographical distribution of dentist is limited. Spatial inequalities in health present significant economic and political challenges for the governments of many countries. Spatial inequality in the form of the excessive concentration of a dentist-to-population ratio in some provinces may impose a problem of oral health on community.
Shortage of dentists is one of the most common health workforce problems around the globe.,, Thailand has various policies to increase the number of dentists for several decades. The total number of dentists has been increasing continuously. In 2015, there were about 4600 dentists working in public sectors under the Ministry of Public Health with the country’s total demand of 7300 dentists. Annually, there have been 500–600 dentists who graduated; however, about 200–300 dentists resign. The country’s dentist-to-population ratio was 1:12,000 against the target of 1:8,000. In addition, the most serious problems are maldistribution of dentists among regions and between rural and urban areas. On comparing the dentist-to-population ratio between regions, the dentist-to-population ratio was found to be 1:10,000 in the Central region whereas it was 1:16,000 in the Northeast region. In some urban settings, the ratio was only 1:5,000 compared to 1:16,000 in the rural areas. This unequal dentist distribution needed proper systematic approaches of allocation improvement. Therefore, the root causes of the problems should be relevantly identified. The most common hypothesized cause of maldistribution is the supply side concerning administrative factors such as the government’s human resource policy. There was no study to identify the demand side concerning the area context factors such as socioeconomic contexts. Therefore, this study aimed to determine spatial relationship and inequalities of dentist distribution in Thailand.
| Materials and Methods|| |
This spatial analysis study was conducted after obtaining approval from the Khon Kaen University Ethics Committee (reference no. 0514.1.27/3519), Khon Kaen, Thailand.
Thailand is a country in Southeast Asia, situated between 15°00′ N latitude and 100°00′ E longitude and occupies an area of 514,000 km2. Myanmar, Cambodia, Laos, and Malaysia share borders with Thailand.
Dentists and population data
The inclusion criteria are only regarding the dentists that work in public sectors. This study retrieved the data of dentists and population in each province between 2007 and 2015 from the link http://bps.moph.go.th/new_bps/healthdata of the Ministry of Public Health, Thailand, for analyses.
The geographic administrative area data of Thailand were retrieved from the DIVA-GIS (http://www.diva-gis.org/gdata), which is publicly available. The provincial maps of Thailand database were processed with DIVA-GIS for further GIS-based analysis to identify spatial distribution of a dentist-to-population ratio by obtaining the province-level polygon map that contains information regarding latitudes and longitudes of each province. Thailand used to have 76 provinces, but a new province, Bung Kan, was established in 2011. However, the data of the new province are incomplete. Therefore, we used the map of 76 provinces for analysis. The dentist-to-population ratio was then paired with the province-level layers of polygon and points by administrative code of each province.
Spatial visualization and analysis
A dentist-to-population ratio was calculated for analysis. The map of dentist-to-population ratio was prepared using the Quantum Geographic Information System (QGIS, version 2.18.10), downloaded from https://www.qgis.org/en/site.
Global and Local Moran’s indices were calculated to determine autocorrelation value and local indicator of spatial autocorrelation (LISA). LISA was analyzed using GeoDa software, version 18.104.22.168, downloaded from https://spatial.uchicago.edu/software. The level of significance was set at P value of <0.05 and simulation run to 999. We used automatic Euclidean weight distance, which matched with the assumption that each province has at least one neighboring province. Interpretation of the LISA significance map includes the following categories: “high–high” indicates a clustering of high-value rates (positive spatial autocorrelation), “low–high” indicates that the low-value rates are adjacent to high-value rates (negative spatial autocorrelation), “low–low” indicates clustering of low-value rates (positive spatial autocorrelation), “high–low” indicates that high-value rates are adjacent to low-value rates (negative spatial autocorrelation), and “not significant” indicates that there is no spatial autocorrelation.
The outcome of Moran’s I identifies the intensity of spatial autocorrelation along with the result of statistically significant test, that is, the P value. The following mathematical representation exhibits the computation of Moran’s I:
where Wij is the spatial weight between a dentist-to-population ratio in provinces i and j; N is the total number of spatial units; S0 is the aggregate of all spatial weights; and xi and xj are the number of a dentist-to-population ratio in provinces i and j, respectively.
| Results|| |
In 2007, the total number of dentists in public sector was 4653 whereas it increased to 6953 in 2015. Between 2007 and 2015, the overall dentist-to-population ratio improved. The average number of a dentist per population improved from 19,534.03 ± 8,428.28 to 11,085.14 ± 3,916.69.
Visualization quintile map of a dentist-to-population ratio in 2007 and 2015 was developed and is presented in [Figure 1]. In 2007, the dentist distribution in the Central and the North regions was better than that of the South, and the worst was found in the Northeast region. There are 13 provinces in the Northeast which show that one dentist was working for as high as 25,333.75–62,863.71 population. The overall distribution improved in 2015 with similar pattern. Some provinces had much better dentist distribution than other provinces such as Chiang Mai, Khon Kaen, and Songkhla [Figure 1].
|Figure 1: Quintiles of provincial dentist-to-population ratio in 2007 and 2015|
Click here to view
Spatial analysis was performed to identify geographic clustering relationship of dentist distribution in Thailand. The global Moran’s I of 2007 and 2015 are presented in [Figure 2], which indicate clustering patterns of dentist distribution in Thailand. However, the clustering relationships were fewer in 2015 when compared to those in 2007 [Figure 2].
|Figure 2: The global Moran’ I of dentist distribution in Thailand, 2007 and 2015|
Click here to view
LISA of dentist distribution in 2007 and 2015 was calculated [Figure 3]. The outputs from LISA represent spatial autocorrelation of a dentist per population at provincial level. This analysis focused on the univariate spatial distribution. In 2005, there were 15 low–low clustering in the Central region and 1 in the North. However, there was no low–low clustering in the South and the Northeast. On the contrary, there were eight high–high clusters in the Northeast region. Only two high–low clusters and one low–high cluster were indicated in the Northeast region. In 2015, there were only four low–low clusters of which two clusters were in the North and one cluster each in the Central and the South regions. Nine high–high clusters were found in the Northeast, which were more than those in 2007. There were two low–high clusters and one high–low cluster identified in the Northeast region.
|Figure 3: Local indicator of spatial autocorrelation (LISA) cluster of a dentist to population in 2007 and 2015|
Click here to view
LISA clusters of gross provincial product (GPP) of the year 2007 and 2015 were calculated [Figure 4]. The spatial autocorrelation of GPP and the dentist-to-population ratio were calculated [Figure 5]. The hotspot (high–high cluster) represented provinces with high GPP and had high dentist-to-population ratio surrounded by provinces with the same pattern. There was no hotspot in 2005 and only one in Chonburi in 2015. The coldspot (low–low cluster) of spatial autocorrelation between the GPP and dentist-to-population ratio represented provinces that had low GPP and had low dentist-to-population ratio around provinces with the same pattern. Most of the low–low dentist-to-population ratio clusters were in the Central region near Bangkok and one each in Mae Hong Son, Lampang, and Kabi. There were 11 low–high clusters in the Northeast region in 2007 and 10 clusters in 2015. It reflected the province with low GPP and high dentist-to-population ratio was surrounded by a province with similar pattern. Eleven high–low clusters were found near Bangkok, which were reduced to only two clusters. This analysis shows association of spatial autocorrelation with GPP and dentist-to-population ratio [Figure 3][Figure 4][Figure 5].
|Figure 4: Local indicator of spatial autocorrelation (LISA) cluster of gross provincial product in 2007 and 2015|
Click here to view
|Figure 5: Bivariate local indicator of spatial autocorrelation (BiLISA) cluster of gross provincial product and a dentist to population in 2007 and 2015|
Click here to view
| Discussion|| |
The results suggest that between 2007 and 2015, the total number of dentists increased. Thailand had both the shortage in absolute number and maldistribution of dentists in public sectors. It is similar to that in other studies, indicating there was maldistribution of both general practitioners (GPs) and specialists., Maldistribution is the one that indicates inequality associated with area. There were several possible reasons for the maldistribution of dentists working in government hospitals.
Thailand has implemented various policies, such as increased number of dental students to be trained in faculties of dentistry nationwide and precontract for dentists to work in rural areas after graduation, to solve the shortage of dentists. The inequality of dentist distribution has improved in the Central, North, and South regions. However, the problem has not much improved in the Northeast region. In 2015, about 9 provinces had the poorest distribution of dentists when compared with 13 provinces in 2007. In addition, there were four provinces (Mukdahan, Loei, Ubon Ratchathani, and Nakhon Ratchasima) in which the distribution was found to be poorer in 2015 than in 2007. The possible explanation in terms of supply side was that health-care facilities are related to the Ministry of Interior’s regional administration where each province has a general hospital (>120–400 beds) or regional hospital in some big province (>500 beds), and each community has one community hospital (10–120 beds). A community hospital had 1–7 dentists, whereas a general hospital had 8–15 dentists and >10 dentists in a regional hospital. Most of the provinces, such as Sing Buri, Angthong, and Ratchaburi, in the Central region were small. Some general hospitals served only 100,000–200,000 population and a community hospital served fewer than 10,000 population. Mae Hong Son province in the North had a low–low dentist-to-population ratio cluster in both 2007 and 2015 because of its less populous nature but had big general and community hospitals. It was similar to that of Lamphun province, which changed to a low–low cluster because of population factors and its borders being shared with Chiang Mai, a big educational and economic city.
However, the situation in the Northeast region was different. The Northeast, which is the biggest region, covers one-third of the country’s land areas and about one-third of population lives in its 20 provinces. Most of hospitals in the Northeast region were community hospitals that had a capacity of 10–120 beds but served more than 100,000 population. In addition, most of general and regional hospitals had a capacity of more than 500,000 and about half served more than 1 million population. Therefore, the implementation of the government policies could improve the dentist-to-population ratio but still not much has been done as compared to other regions. The possible explanation of the low dentist-to-population ratio in Khon Kaen, Chiang Mai, and Songkhla was regarding the demand-side factors. These provinces had dental faculties of well-known universities. There were more dentists who have studied, worked, and taught there. In addition, the rapid economic development, the likelihood has better resulted in these provinces could keep dentists working in this city centers than their rural or neighboring provinces. GPP indicates the provincial economic status. In this study, there was a shortage of dentists in areas with low GPP and there were more dentists in areas with high GPP. The economic status of Khon Kaen as a province was not good but was very good in the city center where the biggest university of the region was located and more dentists practiced and lived there, whereas the socioeconomic condition of its neighboring provinces still did not favor for dentists to live. There these factors explained their low–high clustering relationship. Mukdahan, one of the new provinces in the furthermost of the Northeast region, had changed from low–high in 2005 to high–high clusters of dentist distribution in 2015. The explanation should be that after being a province, Mukdahan community hospital became a general hospital; therefore, there were more dentists allocated and worked there. However, as it has a bridge over the Mekong River to Lao PDR, there was more economic development, and both tourism and international trade acted as pulling factors for more migration into the province. Therefore, the dentist-to-population ratio became poorer, which was similar to that in its neighboring provinces forming a high–high cluster. Chonburi where Pattaya is located is both an industrial and a tourist destination, and has changed from nonsignificant to low–high cluster of a dentist-to-population ratio of which the results of economic development that overcome the effect of migration that the neighboring provinces were facing. Satun province in the South region changed from nonsignificant to low–low cluster of a dentist-to-population ratio was from being small province with continuous economic development from tourism and fishery.
The aforementioned findings were similar to those reported in some studies on distribution of primary health practices both medical and dental. An inequitable spatial distribution of GPs was found in Adelaide Metropolitan of which about 16% residents considered to be living in areas of GP workforce shortage. Residents in the outer suburbs and those with lower social economic status appeared to be the most disadvantaged. Another study found that GPs and family physicians increased in areas gaining specialists, and specialists increased in areas where considerable decreases in GPs and family physicians were observed. Yet another study indicated that there was a negative relationship between private and public practicing dentists within district in Australia, and private dentists were not substitutes for public dentists.
Spatial analysis helps identifying dentist shortage and geographical distribution. The traditional methods use administrative boundaries such as counties as the basic spatial units to calculate dentist-to-population ratios and identify shortage based on those ratios. Such approaches have been criticized for their inability to account for either the spatial variations of population demand and dentist supply within those boundaries or for population–dentist interactions across them. Spatial analysis can assist current and future practicing dentists, dental school administrators, and policymakers in making informed decisions to determine suitable practice locations, dental school admissions criteria, and target areas for public health initiatives. The problem of disparity and inequality of the dentist should use spatial analysis to clarify the problems.
This study included dentists in the public sector alone for the reason of accuracy. Data of dentists in private sector were collected uncover. Future studies are recommended using data of dentists both in the public sector and in the private sector when the process of collecting data has covered.
| Conclusion|| |
Although the number of dentists in Thailand has increased between 2007 and 2015, misdistribution has improved in the North, the Central, and the South regions. In the Northeast region, the dentist-to-population ratios in its provinces have improved but still are the worst in the country with both supply side of resources allocation and demand side of socioeconomic development pushing the dentists out and pulling in the factors for more population to immigrate into their areas. The spatial analysis helps clarify the possible causes of the dentist misdistribution, an inequity.
Financial support and sponsorship
Conflicts of interest
There are no conflicts of interest.
| References|| |
Zhou F, Guo H, Hao Z. Spatial distribution of heavy metals in Hong Kong’s marine sediments and their human impacts: A GIS-based chemometric approach. Mar Pollut Bull 2007;54:1372-84.
Kruger E, Whyman R, Tennant M. High-acuity GIS mapping of private practice dental services in New Zealand: Does service match need? Int Dent J 2012;62:95-9.
Higgs G, Richards W. The use of geographical information systems in examining variations in sociodemographic profiles of dental practice catchments: A case study of a Swansea practice. Prim Dent Care 2002;9:63-9.
Chaikaew N, Tripathi NK, Souris M. Exploring spatial patterns and hotspots of diarrhea in Chiang Mai, Thailand. Int J Health Geogr 2009;8:36.
Hendryx M, Fedorko E, Anesetti-Rothermel A. A geographical information system-based analysis of cancer mortality and population exposure to coal mining activities in West Virginia, United States of America. Geospat Health 2010;4:243-56.
Geraghty EM, Balsbaugh T, Nuovo J, Tandon S. Using geographic information systems (GIS) to assess outcome disparities in patients with type 2 diabetes and hyperlipidemia. J Am Board Fam Med 2010;23:88-96.
Hixon AL, Buenconsejo-Lum LE, Racsa CP. GIS residency footprinting: Analyzing the impact of family medicine graduate medical education in Hawai’i. Hawaii J Med Public Health 2012;71:31-9.
White DA, Anderson RJ, Bradnock G, Gray MM, Jenkins P. The use of a geographical information system in investigating dental services. Community Dent Health 2000;17:79-84.
Kim S. Spatial Inequality and Economic Development: Theories, Facts, and Policies. Commission on Growth and Development Working Paper No. 16. Washington, DC: World Bank; 2008. Available at: https://openknowledge.worldbank.org/handle/10986/28050.
Kurcz R, Kruger E, Tennant M. Using GIS to analyse dental practice distribution in Indiana, USA. Community Dent Health 2013;30:155-60.
Widström E, Tiira H, Tillberg A. Public dental service personnel facing a major health care reform in Finland. BDJ Open 2019;5:5.
Liu DL, Xie YF, Shu R. Statistical analysis of current oral health care and dental education resources in China. Chin J Dent Res 2019;22:37-43.
Pagaiya N, Phanthunane P, Bamrung A, Noree T, Kongweerakul K. Forecasting imbalances of human resources for health in the Thailand health service system: Application of a health demand method. Hum Resour Health 2019;17:4.
Mitchell A. The ESRI Guide to GIS analysis: Geographic Patterns & Relationships. 1st ed. Redlands, CA: Environmental Systems Research Institute Press; 1999. p. 186.
Susi L, Mascarenhas AK. Using a geographical information system to map the distribution of dentists in Ohio. J Am Dent Assoc 2002;133:636-42.
Wright SE. The spatial distribution and geographic analysis of endodontic office locations at the national scale. J Endod 1994;20:500-5.
Gabriel M, Cayetano MH, Galante ML, Carrer FC, Dussault G, Araujo ME. A global overview of the geographical distribution of dentists: A scoping review. JDR Clin Trans Res 2018;3:229-37.
Kruger E, Whyman R, Tennant M. High acuity GIS comparison of dentist and doctor surgery locations in Auckland, New Zealand. Community Dent Health 2013;30:83-7.
Roeger LS, Reed RL, Smith BP. Equity of access in the spatial distribution of GPS within an Australian metropolitan city. Aust J Prim Health 2010;16:284-90.
Krishnan V. A macro model of change in specialty and spatial distribution of physicians in Canada, 1971–1981. Socioecon Plann Sci 1992;26:111-27.
Gächter M, Schwazer P, Theurl E, Winner H. Regional density of private dentists: Empirical evidence from Austria. Community Dent Oral Epidemiol 2014;42:20-9.
Luo W. Using a GIS-based floating catchment method to assess areas with shortage of physicians. Health Place 2004;10:1-11.
Krause D, Frate DA, May WL. Demographics and distribution of dentists in Mississippi: A dental work force study. J Am Dent Assoc 2005;136:668-77.
[Figure 1], [Figure 2], [Figure 3], [Figure 4], [Figure 5]