Dr L Morad, Consultant Psychiatrist, Department of Psychiatry, Waikato Hospital, Hamilton, New Zealand telephone +64 7 839 8603, fax +64 7 839 8737, Email firstname.lastname@example.org
Dr M Morad, Lecturer, Department of Geography, The University of Waikato, Hamilton, New Zealand telephone +64 7 838 4466 x 8429, fax +64 7 838 4633, Email: email@example.com
This research explores the feasibility of projecting demand for psychiatric services in the Waikato Region of New Zealand, using Geographic Information Systems. The latter allow the integration of demographic and spatial data, enabling mental health planners to examine geographical patterns of potential demand, and plan the distribution of psychiatric facilities more efficiently.
The spatial analysis of psychiatric phenomena has been the subject of research by planners and scientists for a long time. In early days, the focus of that attention was on the 'ecological' associations between mental illness and geographical space (Faris and Dunham, 1939). In the 1960s the interest shifted towards mental health reforms, especially in the United States. This was prompted largely by the publication in 1961 of an historic US Congressional document entitled 'Action for Mental Health', which recommended reducing the number of large psychiatric hospitals in favour of smaller community-based facilities. The recommendations were subsequently written into the 1963 Community Mental Health Centers Act (Mechanic, 1969).
The mental health reforms in the US were followed by the publication of important studies on the utilisation and accessibility of psychiatric services (Holton et al, 1973; Smith, 1976; Dear, 1977; Dear 1978). A study by Dear (1977) was probably the first important inquiry into the locational factors in the demand for psychiatric services, which drew on a PhD thesis he had submitted in 1974 to the University of Pennsylvania. That author observed that the demand for psychiatric care could not be measured in the same way as demand for other commercial services, as psychiatric service delivery was not a problem of simply locating the facility close to the population in need. According to Dear, psychiatric clients were generally tolerant of distance, although this 'elasticity' in demand appeared to vary between different social groups. For example, white young middle class individuals seemed more tolerant of distance (perhaps for reasons of anonymity), compared to low-income inner-city residents or members of the black community for whom transport costs were prohibitive.
The relative flexibility of psychiatric clientele over distance considerations has, however, been challenged in other studies. Holton et al (1973) had made the observation that demand for mental health care outstripped supply, and that the psychiatric services were operating in a sellers' market. This situation is, of course, likely to change in future as public health contracts may be awarded to psychiatric-care providers willing to locate close to the centres of demand. Writing in a major publication on measuring mental health needs in community psychiatry, Strathdee and Thornicroft (1992, p140) observed that:
'Services should be local and accessible and to the greatest extent possible delivered in the individual's usual environment... Services should be consumer-orientated, that is, based on the needs of the user rather than those of providers'.
Geographic information systems can play a very useful role in bridging the information gap between health service providers and their client groups, in the post-health reform era (Barnes and Peck, 1994; Hale, 1991). Unlike conventional database packages, geographic information systems provide spatial analytical tools capable of accommodating the 'what if' type of management scenarios. GIS have already been employed to find suitable locations for health facilities, including psychiatric units (O'Hara, 1993; Richie, 1993). They have also been used to solve networking problems and optimise travel times for both medical clients and the health-service providers (Furbee and Spencer, 1993; Norris, 1994).
The GIS aspects of this research were conducted through the medium of Supermap, a demographic GIS system distributed by Statistics New Zealand. The project focused on projecting areas of potentially high demand for psychiatric services, by using demographic and spatial data to point out areas of high vulnerability to psychiatric disorders.
Methods and Aims
An important objective of this research was to explore the use of GIS in planning a psychiatric rehabilitation service for clients suffering from chronic mental illness in the Waikato Region. At this preliminary stage in the GIS application, the main purpose of the project was to use GIS as a spatially referenced database, aiding in gathering, analysing, mapping and querying data on potentially vulnerable areas within the catchment area of Health Waikato. Of course, GIS provides a much wider range of research and planning opportunities than these, but in the authors' opinion, it is important first to highlight the potential of GIS using a modest practical task before recommending wider application. For example, no attempt has been made yet to design a management information system for mental health services, although such a goal could be set at a later stage.
As in all GIS projects, data capture constitutes the most costly and time-consuming phase in starting a GIS project (Aronoff, 1989). It was, therefore, important to decide what the data requirements were, what data where available and how much it would cost to buy or key in the information. In planning for a psychiatric service, the problems of data capture are compounded by the fact that data on mental illness are very scarce, and issues of privacy and confidentiality prevent a liberal approach to data gathering. For example, there are no direct data from population censuses pertaining to mental health, as this could potentially violate the spirit of a free society. The practical and ethical implications of gathering data on mental illness are well documented in the psychiatric literature (Wing et al, 1992).
In the absence of comprehensive census data on psychiatric issues, three alternative data sources have been used in planning psychiatric facilities. These are:
The GIS solution proposed here rests on the latter two methodologies, which may be especially appropriate for community psychiatric purposes. As observed by Craig and Boardman 1991 (p178):
'In most areas, there are no extensive case registers, and no detailed population surveys have been conducted, but recent reports suggest a third possible approach... Geographic Information Systems (GIS) are a new technology... Such data collected at the time of the UK national census... can be used to highlight areas of social deprivation - a factor which is believed to be associated with higher prevalence of psychiatric disorder.'
The GIS-based research of psychiatric health in Waikato began by examining the findings of two recent surveys of mental ill health in the region, conducted by the Mental Health Foundation of Aotearoa/New Zealand (Bridgman 1994). The surveys were especially useful in identifying the socioeconomic and ethnic makeup of the clients of Health Waikato's psychiatric services and other organisations concerned with mental health issues. This insight was important in determining which relevant demographic data to focus on in assessing vulnerability to psychiatric illness (and demand for psychiatric services) in the region.
Three main categories of long-term psychiatric problems were highlighted in the surveys: chronic psychotic, chronic non-psychotic and drugs/alcohol related disorders, according to their socioeconomic backgrounds. The surveys' results were consistent with trends elsewhere, where a relatively high proportion of people suffering from mental illness tend to be female, unemployed and/or living alone. Also, the survey revealed that the distribution of mental illness categories differs among ethnic groups, where Waikato's White community appear more vulnerable to non-psychotic mental illness than the Maori. Nonetheless, Maori psychiatric clients are over-represented (as a proportion of the region's population) in the psychotic and alcohol/drug abuse categories.
Using similar ethnic and socioeconomic indicators as those used in the above surveys, the authors of this paper set out to examine the regional pattern of vulnerability to psychiatric illness and potential demand for psychiatric services. The demographic factors (on gender, ethnicity and socioeconomic conditions) selected for this exercise were included in a spreadsheet. At this preliminary stage in the research, a simple procedure was followed to calculate a standardised score of vulnerability. The original demographic data were deducted from the national average, and the sum of these differentials were used to compute a Zscore (a normal distribution statistic) for each district in the region. The Zscores (which were computed by the STANDARDIZE function in an EXCELL spreadsheet) were finally converted to a percentile scale of vulnerability and demand for psychiatric services.
The substantial analysis was conducted on a PC-based Supermap demographic GIS, with additional analytical work made in EXCELL. In addition to providing demographic data, Supermap can also create maps highlighting legend classes drawn in spatially-correct zones. Other topographic phenomena available in the Supermap package include roads, waterways, lakes, rail links etc. The data could be visualised at different levels of generalisation, with reference to statistical meshblocks (as defined by the Supermap distributors, Statistics New Zealand), local authority boundaries or Area Health Boards. It is also possible to conduct distance and buffer (zoning) exercises to examine, for instance, the employment status of adults living within a kilometre of a particular business centre or along a development corridor (eg a major highway). For advanced AM/FM (automated mapping and facilities management), the data in Supermap could be ported to more powerful GIS platforms such as Arc/Info, Intergraph, Map Info or Genamap which can handle more sophisticated management operations such as network analysis, rapid response assignments and optimising the distribution of clinics, equipment and personnel.
According to the present analysis, the districts of Waitomo and South Waikato, in the centre of the North Island of New Zealand, may be most vulnerable to mental disorders, while the demand base for psychiatric services is likely to be numerically small in the districts of Franklin and Matamata-Piako (respectively, to the north west and north east of Hamilton).
For a closer examination of potential demand for psychiatric services, vulnerability was also calculated for all areas under the jurisdiction of Waikato Area Health Board (in 1991). Huntly West (a depressed town affected by the decline in the mining industry) was found to have the highest vulnerability rating, Acacia Bay (an affluent location the Waikato Region) was least vulnerable, with the middle point occupied by Hamilton Central (Hamilton's main commercial area).
Discussion and conclusions
The research established Tokanui as the area with highest demand for psychiatric services within the Waipa District in the Waikato Region. This finding appears to confirm the authors' view that carefully selected demographic data could be used to project demand for psychiatric services within Health Waikato's territories, as Tokanui is the site of a major psychiatric hospital in the region. It is important, however, to avoid overgeneralisation, as the present findings are yet to be tested (within each district) for the whole of Waikato over several time intervals.
Another inquiry may also be needed in future to compare the current location of psychiatric services in Waikato with areas of projected demand. Of course, location issues are often compounded by many factors other than local demand, and these would have to be considered before deciding on service sites. As mentioned earlier in the introduction, the patterns of access and demand for psychiatric services are distorted by factors relating to the socioeconomic conditions of the area from which clients come, available means of transportation as well as considerations of choice and privacy. The latter appears to influence particularly middle class clients in deciding where and how far to travel to obtain a psychiatric service. Economic considerations on the part of service providers are also important. For instance, it is often cheaper to locate a psychiatric facility in urban centres enjoying high connectivity within the region's transportation network, and where the transportation costs for the psychiatric staff is relatively small.
However, despite the limitations stated above, a number of tentative conclusions about the spatial distribution of psychiatric services in Waikato may be drawn from this research.
- The overall concentration of psychiatric facilities in Hamilton (the capital of the Waikato Region) may only be partially justified (in terms of the economics of health-care provision), but appears to present the bulk of potential psychiatric clients with access problems.
- The planned closure of Tokanui Hospital (scheduled for the end of 1995) may accentuate these difficulties for clients, since the areas with highest potential demand (on a regional scale) are situated south of Hamilton (Tokanui Hospital's catchment area).
- It would also seem that psychiatric services in several parts of the Waikato Region may have to be expanded in future to reach out to a substantial client group affected by the high costs of travelling to and from Hamilton to obtain service.
- Accessibility of services in areas with a high proportion of Maori clientele can be maximised by providing a culturally acceptable service. The proportion of Maori population in Waikato is generally higher than the national average.
The proliferation of Geographic Information Systems in public and commercial organisations is likely to grow, as these organisation strive to maximise the quality of their services and the efficiency of their delivery. The health services in most developed countries already make up an important proportion of GIS user groups, although New Zealand's health organisations are yet to make a substantial move in this direction. An important aim of this research was to provide a tangible example of the benefits to psychiatric-health specialists of using GIS for planning and monitoring a service, especially for community-based psychiatric rehabilitation.
The authors acknowledge gratefully help received from Health Waikato, and also from The University of Waikato, during the preparation of this research.
Return to Contents
Aronoff, S (1989), Geographic Information Systems: A Management Perspective. Ottawa: WDL Publications.
Barnes, S and A Peck (1994), 'Mapping the Future of Health Care: GIS Applications in Health Care Analysis,' Geo Info Systems, 4(4):30-39.
Bridgman, G (1994) Two Surveys of the Prevalence of Mental Ill Health in the Community - Summary Document; New Zealand: Mental Health Foundation of Aotearoa/New Zealand.
Clifford, P, A Charman, Y Webb, TJK Craig, and D Cowan (1991), 'Planning for community care: The Community Placement Questionnaire', The British journal of clinical psychology, 30 (3).
Craig, T K J and A P Boardman (1991) 'The contribution of epidemiological research to community psychiatry', Community Psychiatry. London: Churchill Livingstone.
Dear, M (1977), 'Locational Factors in the Demand for Mental Health Care', Economic Geography, 53(223-240).
Dear M (1978), 'Planning for Mental Health Care - Reconsideration of Public Facility Location Theory' International Regional Science Review, 3(93-111).
Faris, D E L and H W Dunham (1939), Mental Disorder in Urban Areas. Chicago: Hafner / Chicago University Press.
Fryers and Greatorex (1992) ' Case registers and mental health information systems', in Thornicroft, G, C R Brewin and J Wing (eds) Measuring Mental Health Needs, London The Royal College of Psychiatrists / Gaskell.
Furbee, P and J Spencer (1993), 'Using GIS to Determine Travel Times to Hospitals', Geo Info Systems, 3(8): 30-34.
Hale, D (1991) 'The Heath Care Industry and Geographic Information Systems', Mapping Awareness, 5(8): 36-39.
Holton, W E B and B M Krame (1973), 'Locational Process: Guidelines for Locating Mental Health Services', Community Mental Health Journal , 9:270-280.
Mechanic, D (1969), Mental Health and Social Policy. New Jersey: Prentice Hall.
Miller, G H, Dear, M and D L Streiner (1986), 'A Model for Predicting the Utilization of Psychiatric Facilities', Canadian Journal of Psychiatry, 31: 424-430.
Norris, J (1994) 'Managing Healthcare Networks- It's Geographic', Business Geographics 2(4): 34-35.
O'Hara, V, (1993) 'Health Care Professionals Find the 'Right' Location', Business Geographics 1(5): 40-41.
Richie, W (1993) 'Desktop Mapping Helps Rural Wales Target Health Resources' GIS Europe 2(10):50-51.
Royal College of Psychiatrists (1988) Psychiatric Beds and Resources - Factors Influencing Bed Use and Planning. London: Royal College of Psychiatrists / Gaskell.
Smith, C J (1976), ' Distance and Location of Community Mental Health Facilities: A Divergent View Point', Economic Geography, 52: 181-191.
Strathdee, G and G Thronicroft (1992), 'Community sectors for needs-led mental health services', in Thornicroft, G, C R Brewin and J Wing (eds) Measuring Mental Health Needs, London The Royal College of Psychiatrists / Gaskell.
Thornicroft, G (1991) 'Social Deprivation and Rates of Treated Mental Disorder', British Journal of Psychiatry, 158: 475-484.
Wing, J, C R Brewin and G Thornicroft eds (1992), 'Measuring and meeting mental Health needs', in Thornicroft, G, C R Brewin and J Wing (eds) Measuring Mental Health Needs, London The Royal College of Psychiatrists / Gaskell.
All pages copyright ©Priory Lodge Education Ltd 1994-1999.