Application of GIS in health care facilities planning

Application of GIS in health care facilities planning

Abdul Kader A. Murad
Department of Urban & Regional Planning, Faculty of Environmental Design
P.O.Box 80210, Jeddah-21589, Saudi Arabia
amurad25@hotmail.com

Health care facilities planing is one of the planning fields that uses spatial data in its resources allocation process. Health authorities and officers have been required to keep registers and carry out analysis of costs, benefits and health needs. This is quite a complex procedure as the information needed by the health authorities is extremely varied. Today, Geographical Information Systems (GIS) provide useful techniques regarding capturing, maintaining and analyzing the spatial data. The aim of this research is to identify the ways in which GIS can be applied to help planners in planning and monitoring the location of Health care facilities.

Healthcare Planning and GIS
One of the basic objectives of healthcare Planning in any part of the world is to have an equal access to health care for all, it respective of ability to pay. This means that every residence should have equal chance to go to clinics and hospitals. To meet this objective and other ones, health authorities are required to make careful analysis about the real demand and supply of health care facilities at their areas. These analysis and studies can be grouped into three main areas, which are a) spatial changes in health status, b) spatial epidemiology, and c) health care facilities accessibility and utilization. Each one of these topics has a spatial dimension, which means that GIS can be used for their studies. The next part will describe more about each group and illustrate the possible uses of GIS is them.

Spatial Changes in Health Status
One of the facts about health status in a micro/macro scale is that it changes across the space. Health authorities always investigate and analysis the health status at their areas and make sure that health needs satisfied.

Locality definition is considered as an important issue for health care facilities planning studies. The idea here is to determine the socio- economic classification for the area surrounding certain health facilities and then relating the local profiles of such an area with the health care needs. Once the socio-economic status of any location is defined, then GIS can be used to map and tabulate the distribution of such status. A good example of using GIS for linking social profiles with health needs is found by Hirschfield et al., 1995, which have produced patient profiles for an every health facility catchment area. Such studies usually involve matching point-referenced, post coded health data with area socioeconomic data, particularly deprivation indicators (Gatrell and Senior, 1999).

For example, Health status can be viewed through comparing the actual number of moralities in an area with the national average, taking into account age and sex variations in the area concerned (Birkin et al., 1996). Here GIS can be used successfully for describing spatial variation of mortality at parts of any country. Once the mortality rate of each region is entered into the GIS, then the mapping and analysis tools of GIS can be applied to present out the regions that have high rates of mortality. The regions with high rates require more attention from health authorities in order to improve their existing health status. In addition to mortality, there is much other health status indicators that are used by health authorities, such as fertility rates, which help them to assess and monitor the required health services.

Spatial Epidemiology
The second area of health care research is known as spatial epidemiology. There are several questions that are commonly asked in spatial epidemiology studies, which include: where are the incidences located? what are the environmental characteristics of these areas?, are there associations between health incidences and the environment at other locations?, and, what patterns are evolving? (Nicol, 1991). GIS is considered as a useful tool for answering the preceding questions. For example, GIS can define the actual location of health events, and then overlay analysis can be used to create new spatial relationships and to tag the various socio-economic and environmental information to the health data.

There are several studies that have applied GIS for these issues. For instant Brown et al, 1991, have used GIS for the mapping of spatial variations in health care provision in Merseyside, Wrigley, 1991, have also used GIS in mapping incidence diseases in relation to population types. Another example of GIS application in epidemiology is called the Health and Environment Geographical Information System (HEGIS), which is being established in Europe by the world health organization (Nicol, 1991). It involves the creation of European wide environment data set, and the aim is to research relationships between health and the environment, to aid policies and management (ibid.). Most of spatial epidemiology studies must be based upon accurate knowledge of the population. Therefore, access to details of population composition and socio-economic characteristics are very necessary for these studies.

Spatial epidemiology studies are concerned with finding good description of spatial incidence of diseases as well as the modeling of such incidence. One way of describing the spatial distribution of a certain disease is by visualizing the GIS choropleth maps that show the spatial distribution of such disease. In such maps, disease rates are plotted over the base map to define the areas that are highly affected from the related disease. Further analysis and modeling of the spatial incidence of diseases can be carried out using for example Kernal or density estimation technique that is used to predicting the spatial variation in diseases risk. (Gatrell and Senior, 1999).

Healthcare Facilities Accessibility and Utilization
This area of health care research concerns with all the issues that are related to the location of the health care facilities. These issues include the optimal location of hospitals and clinics, the relationship between existing locations and health care needs, and the assessment of facilities accessibility. The planning of any health care system should have answers to the following important questions: What are the population needs for health care and how should resources be allocated to the population they are designed to serve? (Gatrell and Senior, 1999).

GIS can assist in finding comprehensive answers for the proceeding questions. For example, Jonhs and Bentham 1995 have used GIS to test for a relationship between health outcomes and accessibility. Forbers and Todd, 1995, have also used GIS to evaluate the potential locations for a new radiotherapy unit for cancer treatment in northwest England.

Finding the best location for a health care facility is considered as one of the health authority task in order to optimize resources. The best location for a facility can be identified using for example location - allocation models that are now been integrated to the modern GIS softwares such as Arcinfo Version 7.

Evaluating the accessibility of existing health facilities is also another task carried by health authorities. Hear the health planners determine the areas, which have poor accessibility for certain, health facility and then prepare proposals for improving such low level of health care accessibility. One way of improving such accessibility is by increasing the capacity of the related health facility, but this can only be achieved after intensive analysis of the area that contain such a facility.

Health Care Planning Data
In order to evaluate the existing location of a health service or to prepare a new site location health care, planners have to collect large set of data that then can be used for the relevant Health care issue. There are several ways of classifying this type of data, but one method is based on the GIS data format types, i.e. dividing the health data into three main groups which are point, area and line data.

In a GIS, point data can be a model for the locations of residences, hospitals or ambulance stations. Each one of the point data can have different attribute information. For examples the locations of individuals might include attributes for the presence or absence of a disease, age, gender, occupation, and so on.

Population zones or census data on the other hand are examples of area data format which can be captured and stored in GIS and can be used for different health care studies. For instance census data can be used to analyze the socioeconomic status of certain patient locations.

Finally, line data such as road network are used in GIS to study the travel journey to and from health facilities locations. They are also used to analyze the routes used by emergency vehicles and to identify how fast do ambulance vehicles reaches to patient locations.

Using GIS in Hospitals Facilities Planning
One area of GIS research in health care planning field is concerns with measuring accessibility to health facilities (Section 2.3). This part will concentrate more on this issue and explain how GIS is used to analyze accessibility to hospitals in Jeddah City, Saudi Arabia. There are three important factors that affect the level of accessibility in any facility location. These are the capacity of the facility, the amount of demand for such facility, and the transportation network that communicate such demand to the relevant capacity. All of these three factors are included in the example of Jeddah hospitals.

The Data Base
In order to build this application, the present study has collected large set of data that are then entered into the GIS to form the database of this application. This means that all of the collected data were in a paper format, i.e. they were not digital. Accordingly, all of the collected data have been entered into the GIS using the digitizing method. The present study has captured three major maps (GIS Coverages) and then added to them their relevant attribute (non-spatial) data. These coverages are the road network, the hospital locations and the population coverage. (Figures 1 and 2). The nonspatial data which are linked to those coverages include the hospitals size (capacity), no of people lives in each district of the city, and the population density of these district. All of these data are then used for the modeling process of hospital accessibility.

The Modeling Process
Once the required data have been captured into the GIS, then the following step was to decide about the relevant analysis methods that may be used for determining hospital accessibility. There are three main GIS spatial analysis functions that could be used for analyzing the accessibility of any selected facility location. The first one is known as buffer analysis which draw buffer around existing facilities proportional to the latter.s size and capacity (Davis, 1996). The second method is related to the GIS network analysis module where population in a network is allocated to the nearest facility locations. Site suitability analysis is the third GIS function that identifies sites according to the suitability for the location of facility under a set of certain (ibid.). Looking at these three GIS functions, the present study has selected the second method that is the network analysis module as the analysis tool for determining hospital accessibility in Jeddah. The main reason for this selection is that hospital demand can be included in the network analysis functions while it is not possible to do so by using buffer or suitability analysis functions. In most GIS softwares, the network analysis module consists of several modeling functions including finding shortest path, service area a model, allocate model, location - allocation model and spatial interaction model. The present study has used the service area model that is one of Arcview (a GIS software produce by ESRI, USA) network analysis functions for evaluating hospital accessibility. This function can find the accessible streets within certain distances of a site and accordingly the streets which are not selected by this function are representing the problem areas i.e. they have poor accessibility (ESRI, 1997).

Results and Discussion
The Ministry of Municipal and Rural Affairs (MOMRA) of Saudi Arabia have defined a set of criteria for the planning of hospital locations. These criteria are selected by the present study and applied on every hospital in Jeddah in order to evaluate their locations and accessibility. Public (governmental) hospitals are selected by this study as the main provider for hospital care in Jeddah. In another wards, private hospitals are not included by this study because these hospitals have different planning and management process that can be analyzed by separate studies.

The MOMRA criteria for hospital planning indicate that each service should cover a 4 km catchment area with a population capacity of 60,000 or 2.5 beds per 1000 persons (MOMRA, 1980). These criteria are applied on every hospital using the service area function of Arcview software. The output of the function is a network coverage showing the parts of the city that are within 4 Km at catchment area (Figures 3 and 4).

It is clear from this figure that the northern part of the city is located outside the hospital catchment area. This means that the existing hospitals are not serving the whole city, and there is a large part of the city, which is based on MOMRA criteria, is not within the hospital catchment areas. The next step of the analysis was to define the size of demand within each hospital catchment area. This step will identity how many people be served by the existing hospitals, and the no. of beds that should be provided by each hospital. Relating hospitals demand (D) with hospital capacities is calculated using the following formula

D = P x a

Where P = the total no. Of people living inside hospital catchment, and a = 2.5 / 1000 person

Table 1. Results of 4Km Hospitals catchment areas.

Hospital Name	Hospital capacity (Beds)	Catchment area	Hospital Demand (Beds)
King Fahad	1001	4 Km2	974.5
King Abdulaziz	442	4 Km2	1153.8
Al-Thugar	143	4 Km2	1673.0
King Abdulaziz University	294	4 Km2	1607.9
Total	1880	16 Km2	5410.2

Using Arcview select by them function, the present study has defined the size of people that lives inside each hospital catchment area. Theoretically, all of the people living inside these catchment areas should have free access to hospital care, but practically this is judged by the capacity of each hospital. Accordingly, if the size of demand of any hospital catchment area is more than the hospital capacity then the problem of service shortage will appear. Table 1 shows the results of this application, which are mainly as following:

Existing hospitals capacity are greater than MOMRA criteria e.g. King Fahad Hospital is serving 389815 people where the criteria define hospital size by 60,000 people.
The size of demand within each hospital catchment area is more than hospital capacity. For example, Kin Abdulaziz University Hospital capacity is 294 beds while its demand is 1608 beds.

Accordingly, it can be said that there is a great need to increase existing hospital capacities in order to overcome their existing supply shortages. The resulting shortages also clarify, why private hospitals are distributed at all parts of Jeddah City.

Conclusion
Identifying health care needs is one of the important tasks which health authorities frequently do. Health planners keep analyzing the changes that occur on health care demand. These changes usually require quick respond from health authorities, in order to keep matching health care facilities supply with the relevant demand. This is a large task that needs huge set of data including health facilities location, catchment areas, population statistics, etc.

Information technology in general and GIS in particular can help the health authorities to make their required jobs in an efficient manner. For example GIS can be used to support health planners in reaching a decision regarding increasing hospitals capacities. The present study has illustrated how GIS has been used for Jeddah health care facilities planning. This application has identified the parts of the city, which require more attention regarding their health care supply. Using this application, health planners can allocate areas of the city which having poor accessibility to hospitals and according, a decision regarding improving hospital accessibility can be easily and quickly reached.

References

BaIraqi N., 1997, Demographic changes in Jeddah, unpublished Ph.D. thesis, Jeddah.
Birkin, M., Clarke, G., Clarke, M. And Wilson A., 1996, Intelligent GIS, Geoinformation Int., Cambridge.
Brown P., Hirschfield A, and Batey P., 1991, Applications of geodemographic methods in the analysis of health condition incidence data, Papers in Regional Science, vol. 70, pp 329-44.
Davis B., 1996, GIS a visual approach, Onword Press, Santa Fe. ESRI, 1997, Getting to know Arcview, Geoinformation Int., Cambridge
Forbers H; and Todd P; 1995, Review of cancer services,. Northwest regional health authority, urban research and policy evaluation regional research laboratory, university of Liverpool.
Gatrell A: and Senior M., 1991, Health and Health care application in Longley P, Goodchild M, Maguire D and Rhind D, Geographical information system principles and application, John Wily & Sons, U.K., pp. 925-938.
Hirschfield A, Brown P; and Bundred P; 1995, the spatial analysis of community health services on wirral using Geographical Information System, Journal of the International research society, vol. 46, pp. 147-59.
Johnes A and Bentham G; 1995, Emergency medical service accessibility and outcome from road traffic accidents, Public health Vol, 109, pp. 169-77.
MOMRA, 1980, Planning Criteria for Health Care Facilities, MOMRA, Riyadh.
Nicol J; 1991, Geographic Information System within the national Healths, the scope for implementations, planning out look, vol. 34, No. 1, pp. 37-42.
Wrigley N., 1991, Market-based systems of health-care provision, the NHS Bill, and geographical information system, Environment and Planning A, vol. 23, pp 5-8.