Health GIS - Enabling Health Spatially

2. METHODS The susceptibility of individuals, groups and populations is however related to both time and 2.1 Hazard Identification scale. Briggs (2005) argues that assessment of exposure to urban pollution must balance The first step in a HIA is to identify the “nighttime geography” (where people reside) pollutants that are a risk for human health. Car with the “daytime geography” of pollution. This traffic is a great contributor to particulate matter is of special interest when assessing health (PM) en nitrogen (di)oxides (NOX). Both are effects of traffic related air pollution. However identified to have negative effects on health epidemiologic research restricts the use of (Pope & Dockery, 2006; Samoli et al. 2006). individual exposure estimates in health impact Ozone as a secondary pollutant, formed by assessment. Movement of people can thus not be precursors such as nitrogen dioxides does also taken into account, as well as possible peak cause inflammatory reactions in the lung tissue concentrations in the vicinity of the streets, both (Levy et al., 2001). Finally benzene is added the highly related to traffic. list, as it has a permissible volume up to 5% in petrol and is a causative agent in the 2.3 Exposure Assessment development of leukaemia (Van Larebeke & De Bont, 2006). With the pollutants of interest 2.3.1 Concentration data: Different models identified, it is possible to look at epidemiologic exist to assess traffic-related exposure. Activity- research that establishes the link between an based modelling (AB-models) is seen as a new exposure and the health outcome. technique improving the accuracy of exposure estimates (Janssen & Sanderson, 2004). AB- 2.2 Exposure-Response Evaluation models provide detailed Origin-Destination (O/D) matrices, by estimating activity-travel Key in performing in a health impact assessment schedules for every individual or household. is the concentration-response function (CRF), These schedules are a sequence of travels which links the exposure estimates with the generated by the derived travel demand from the current health burden to estimate an expected activities performed throughout the day. The health response. These concentration-response model can thus predict which activities are functions are derived from epidemiological carried out, where, with whom, when, how long, studies. However the use of these CRF’s requires with which transportation mode and finally compatibility between exposure data used in the which route is followed (Beckx et al, 2008a). By original epidemiological studies and the changing some parameters in the model, this population exposure distribution in the HIA information can be calculated for different where they are applied. When the CRF from the situations. original population is transposed to an other When this extensive information is applied to a study population, it is necessary to characterize transportation network, detailed information the exposure in the target population to mirror as about activity-related traffic flows on the closely as possible the exposure from the original different roads (e.g. average speed and amount of study, e.g. the difference between placements of traffic) and total distances for each trip can be monitor and in the determinants of population calculated. Based on pollutant-specific emission average exposure (i.e. time outdoors, use of air factors, these activity-related traffic flows on the conditioning, exercise and work habits) (WHO, different traffic links can then be converted into 2001). vehicle exhaust emissions. When emissions from other sources, such as industry and agriculture Population exposure can be calculated based on are added and then combined with dispersion the place of residence or on personal exposure. models, these models provide more accurate In most epidemiologic research studies from estimates on hourly ambient pollutant which the functions are derived, the correlation concentrations (Beckx et al, 2008b). of health outcomes with ambient average level of A spatial resolution of 1x1km, on which the air pollution is given, rather than with personal concentration model provides the different exposures. In view of this, place of residence has concentrations, is doesn’t allow to assess the to be selected as the criterion for calculating within-city variability of pollutants, which is population exposure, and the concentration map especially relevant for traffic-related air has to represent the urban background conditions pollution, where street canyons and busy roads typical of the residence areas of the population can lead to local hot spots of air pollution and the (Künzli, 1999). 2

composition of the pollution mixture can differ The census area units Table 1 NO2-concentration per CAU greatly from other locations (Janssen & Sanderson, 2004). However current GIS- are intersected with approaches can offer valuable improvements in acquiring detailed exposure estimates in health the concentration polygons in ArcMap 8.2 studies, by combining the urban background concentrations with detailed population maps (ESRI, Redlands, CA), as can be seen in Figure (Janssen & Sanderson, 2004). 1. These intersection polygons each have a 2.3.2 Exposure estimates: Exposure estimates are obtained by combining population unique combination of a concentration polygon data (place of residence) with concentration estimates. Concentration estimates are provided and a census area unit. Postal address points are in a point-based grid coverage (X-Y coordinates) and need to be converted into 1 km by 1 km added to these unique combination intersections polygon grid coverage (Thiessen polygons) using ArcView GIS 3.2 (ESRI, Redlands, CA) to be and The sum of address points per unique able to combine the two datasets (e.g. Scoggins et al., 2004). Population data is provided per intersection is calculated and the proportion of census area unit (CAU). A census area unit in Flanders has an average population of 666 number of points within an intersection and the people and an average surface area of 1,42 km2. Dasymetric mapping is used to recast the total census area unit is derived. population to every concentration polygon. Dasymetric mapping is a method of areal Figure 1 Population distribution in five census area interpolation enhanced by incorporating ancillary units with address points within the NO2 pollution data. Areal interpolation is a set of techniques to estimate the distribution of a phenomenon grids (population number) across one set of spatial This proportion is used to redistribute the total units called, source units (census tracts), in terms of a second set of spatial units, called target units number of population of a census area unit (the concentration polygons) (Cromley & according to the different concentration McLafferty, 2002). In this study population polygons. numbers are redistributed within a census area Table 1 gives an example of the distribution of unit through the use of postal address points the population of the five census area units (CRAB-points). These postal address points within different concentration polygons within. provide additional information on the (possible) This distribution table can be used to calculate distribution of the aggregated population data the Population Weighted Average, which is within the unit and thus provide a more realistic directly applicable in a health impact assessment estimate of population distribution than assuming and takes into account the distribution of the population is uniformly spread within the population numbers and different concentrations census area (Langford et al, 2008; Briggs et al, within a defined geographic area (Filliger et al., 2007b). 1999). 3

Conci  popi issues a higher level such as the municipality level is adopted, with age and gender PWA  i popi classifications aggregated on such a level that every possible link with individual health records i is made impossible. Nonetheless does the municipality level remain a highly detailed level With PWA = Population weighted average for acquiring mortality data. For data on hospital admissions only the district level is possible for  Conci = Concentration in celli the simple reason that not every municipality has popi = Population in celli a hospital. As close this may resemble to reality, many Next to the fact that health data on a detailed level are very difficult to obtain, two challenges uncertainties however remain. The postal address remain in dealing with health effects on such a small scale. Both have an influence on deciding points make no difference between residential, which level should be the master geography. The first is the modifiable areal unit problem industrial, governmental or other properties. In (MAUP), more specific the scale effect, which states that different statistical results can be this way population numbers are redistributed to obtained from the same set of data when the information is grouped at different levels of industrial areas and office-buildings as well, spatial resolution (e.g. census areas, municipalities, districts, regions) (Wrigley et al., which does not correspond with reality. Counts 1996; Openshaw, 1984). It may be possible that large differences in exposure and health are also not stratified by age or gender, which outcomes between municipalities disappear in further calculations on a higher level, e.g. district makes estimations of specific subpopulations, level. The second issue is the small numbers problem. such as children or elderly, to remain Small-area health data can pose large statistical problems, such as when a difference of one or problematic. The assumption is made that gender two cases can make a huge difference in incidence or prevalence rates (Cromley & and age distribution is spread evenly within McLafferty, 2002). This is highly relevant as this study assesses cause-specific incidence rates every address point. These uncertainties become divided by age and gender. In this way it can be possible that a small municipality only has one or increasingly important as environmental health few deaths from e.g. a respiratory disease at a specific age and thus giving unstable incidence/ studies put more and more emphasis on the mortality rates effects of exposure on critical life stages such as The restrictions from the health data determine to a large extent which zone design is chosen as the early life years and the elderly (Briggs et al, master geography. As it is relevant for policy to have insight on exposure estimates on a very 2007a). detailed scale, this is less the case for health effects. Health effects will therefore be The Population Weighted Average can be calculated on the municipal level for mortality and district level for hospital admissions. For calculated for every geographical scale needed. communication goals however a larger level is more appropriate, such as the district level This resolves the problem of recasting data from (Elliott & Wartenberg, 2004), unless remarkable differences should disappear as discussed in de one spatial resolution to another, but does not MAUP. give directions into the decision which 3. CONCLUSION geographical level should be regarded as the 4 master geography. The geographical level on which health data are aggregated will be decisive on this. 2.4 Health Impact Characterization More and more GIS-applications are being used in HIA’s to integrate the different datasets (Geelen et al., 2009; Mindell & Barrowcliffe, 2005). However there are some issues to bear in mind when integrating all these data, especially regarding health data. In general, a major issue with environmental justice and health research is the difficulty in obtaining data at a resolution and accuracy level sufficient to reliably demonstrate the connections between environmental conditions and health outcomes (Maantay, 2007). As exposure estimates are given on a spatial resolution smaller than the census area unit an equally detailed health dataset should be recommendable. However due to confidentially

New developments in GIS offer useful tools for Dumont, G., Fierens, F., Torfs, R., et al. (2005), dealing with the integration of the different data- Luchtverontreiniging en verkeer. Welke rol speelt sets. They however bring specific assumptions verkeer in de stof- en ozonproblematiek? with them, that, when performing a health impact Milieurapport Vlaanderen MIRA-T 2005. Lannoo: assessment with high spatial exposure and health Leuven, 115 – 128. data resolutions, first need a thorough reflection. Two main issues relevant to health studies are Elliott, P. & Wartenberg, D. (2004), Spatial discussed in this paper. The answer to which Epidemiology: Current Approaches and Future geographical level to use as the master Challenges. Environmental Health Perspectives, geography is not so straightforward. High spatial 112, 9, 998-1006. resolutions give detailed, but unstable results, while aggregated results can become irrelevant Fierens, F. (2008), Dalende NOx emissies - for policy makers as they do not know where to stagnerende NO2 concentraties in stedelijke put priority. Therefore highly detailed exposure omgeving: wat is er aan de hand? IRCEL-VMM, 9. estimates are proposed, to give policy makers insight on where problems are most urgent, but Filliger, P., Puybonnieux-Texier, V. & Schneider, J. for health outcomes, more aggregated zone are (1999), PM10 Population Exposure. Technical preferred to avoid unstable results. Report on Air Pollution. WHO Ministerial Deciding on a master geography also imposes Conference for Environment and Health, London, methods on how to recast data to this level. This 80. is only necessary for population and exposure data. Although dasymetric mapping offers a Forsberg, B., Hansson, H-C. Johansson, C. et al. realistic distribution of the population, (2005), Comparative Health Impact Assessment of uncertainties remain on gender and age local and Regional Particulate Air Pollutants in distribution. These uncertainties can however not Scandinavia. Ambio, 34, 1, 11-19. be resolved without conflicting confidentiality issues. Geelen, L., Huijbregts, M., den Hollander, H., et al. (2009), Confronting environmental pressure, REFERENCES environmental quality and human health impact indicators of priority air emissions. Atmospheric Beale L., Abellan, J., Hodgson, S. & Jarup L. (2008), Environment, 43, 1613–1621 Methodologic Issues and Approaches to Spatial Epidemiology. Environmental Health Perspectives, Janssen, N. & Sanderson, E., ed. (2004), Air pollution 116, 8, 1105-1110. and the risks to human health – Exposure Assessment. Institute for Risk Assessment Sciences, Beckx, C., Arentze, T., Int Panis, L. et al. (2008a), University of Utrecht: Utrecht, 91. Assessing Activity-Related Vehicle Emissions (http://airnet.iras.uu.nl/products/pdf/airnet_wg1_exp through an Integrated Activity-Based Modelling osure_report.pdf) Framework. International Congress on Environmental Modelling and Software, 1177-1186. Künzli, N., Kaiser, R., Medina, S., et al. (2000), Public-health impact of outdoor and traffic-relayed Beckx, C., Int Panis, L., Arentze, T. et al. (2008b), A air pollution: a European assessment. The Lancet, dynamic activity-based modelling approach to 356, 795-801. evaluate exposure to air pollution: methods and application to a Dutch urban area. Environ Impact Langford, M., Higgs, G., Radcliffe, J. et al. (2008), Assess Rev, doi:10.1016/j.eiar.2008.10. 001. Urban population distribution models and service accessibility estimation. Computers, Environment Briggs, D., Fecht, D. & de Hoogh, K., (2007a) Census and Urban Systems, 32, 66-80. data issues for epidemiology and health risk assessment: experiences from the Small Area Health Levy, J., Carrothers, T., Tuomisto, J. et al. (2001), Statistics Unit. J. R. Statist. Soc. A, 170, Part 2, 355- Assessing the Public Health Benefits of Reduced 378. Ozone Concentrations. Environmental Health Perspectives, 109, 12, pp. 9-20. Briggs, D. Gulliver, J. Fecht, D. et al. (2007b). Dasymetric modelling of small-area population Maantay, J. (2007), Asthma and air pollution in the distribution using land cover and light emissions Bronx: Methodological and data considerations in data. Remote Sensing of Environment, 108 (Issue 4), using GIS for environmental justice and health 451-466. research. Health & Place, 13, 32–56. Cromley, E.K. & McLafferty S. (2002), GIS and Mindell, J. & Barrowcliffe, R. (2005), Linking Public Health; The Guilford Press, New environmental effect to health impacts: a computer York/London, 340. modelling approach for air pollution. J Epidemiol Community Health, 59, 1092-1098. Openshaw, J. (1984), The modifiable areal unit problem. Concepts and Techniques in Modern Geography 38. Norwich, UK, Geobooks. Pope, C.A. III & Dockery, D.W. (2006), Health Effects of Fine Particulate Air Pollution: Lines that Connect. J Air & Waste Manage Assoc, 56, 709- 742. Samoli, E., Aga, E., Touloumi, G. et al. (2006), Short- term effects of nitrogen dioxide on mortality: an analysis within the APHEA project. Eur Resp 5

Journal, 27, 6, 1129-1137. Scoggins, A., Kjellstrom, T., Fisher, G., et al. (2004), Spatial analysis of annual air pollution exposure and mortality. Science of the Total Environment, 321, 71-85. Van Larebeke, N. & De Bont, R. (2006), Gezondheidseffecten van benzeen. Rapporten van het Steunpunt Milieu en Gezondheid, 10, 2006. 6

GIS APPLICATION IN A SOCIAL GERONTOLOGY STUDY IN BAGUIO CITY, PHILIPPINES Alicia G. Follosco Cordillera Studies Center, University of the Philippines Baguio Governor Pack Road, Baguio City, 2600 Philippines [email protected] ABSTRACT: This paper comes from a bigger case study of marginalized elderly men and women who are practically living on their own in Baguio City, northern Philippines. The study itself is exploratory on two counts; one, it is a first attempt to describe the strategies employed by this particular group of elderly in coping with the demands of their everyday life in an urban environment and two, which is the focus of this paper, it is an attempt to use GIS to integrate and vi - sualize data generated from key informant interviews. In particular, the social and physical networks of the case study participants were mapped and analyzed. This paper demonstrates that GIS can complement the study of a so - cial issue, like those addressed in social gerontology. It ends with a recommendation that in studying the elderly and in the broader matter of addressing their needs, the use of non-conventional methods, e.g. GIS, could enrich analysis and program planning and implementation. KEY WORDS: elderly, case study, GIS application 1. INTRODUCTION Meanwhile, the application of Geographic Information Systems (GIS) in the health sector is also The study of the elderly population is on the rise glob- taking place worldwide. The realization that GIS is a ally. This sector of the population is particularly in- powerful tool for data integration and spatial teresting to study, not only because they are among visualization came over a decade ago. Since then, those labeled as ‘special groups’, but also because countless applications have been tried and they are the group to which in due time most people implemented, and innovations continue to emerge will belong (Tinker 1981). Also, studies in gerontol- from the public health sector, policy and research ogy are relevant especially as almost all nations are organizations, health care providers and other private now dealing with various demographic developments, sector companies. among them the phenomenon of population ageing – a process by which older persons assume a propor- Health care providers were among the early adopters tionately larger share of the total population (Guan of GIS. They applied the technology in various areas, 2008). In Asia, for example, Natividad (2008) asserts like health policy research, site selection, e.g. that population ageing “is occurring at a much more establishment of service areas and districts, tracking rapid pace than has been experienced in Europe and cases, developing emergency response networks, and North America.” mapping the causes and spread of infectious diseases and modelling how these may be tracked and Among the drivers of the trend in population ageing prevented. Later, the private sector in the health care are declining fertility rates and to a lesser extent, mor- industry incorporated it in marketing studies and tality rate decline, thus rising life expectancies. With business management, among others. the expectation that the number of elderly will in- crease in the coming years and taking into account Human service agencies have also realized that for that societal changes are also occurring especially in them to generate sufficient financial support and to be the traditional practice of care giving where family or able to make sound policy decisions on health and kin groups provide social support in old age, it is human service needs of communities, better timely to address the issue on who will provide sup- information management is necessary. Considering port for the growing number of elderly (Natividad the spatial dimension of their services, displaying 2008). These developments, among others, are fuel- these information through maps is a vital information ing the impetus for research on social gerontology, a management strategy, and GIS has effectively field of gerontology that focuses on how the elderly demonstrated this. However, the application of GIS population and the diversity of their ageing experi- as a complementary tool in qualitative social science ences affect and are affected by social structures. research is not widespread. In the Philippines, studies on the elderly is still limited, more so on the use of GIS for spatial analysis and visualization.

Author’s Guidelines – HealthGIS 2009 It is within the spirit of looking for ways to use GIS in 2. RESEARCH DESIGN social research that this paper was developed. The primary aim is to apply GIS to complement the find- The bigger research combined qualitative and ings and enrich the analysis of the study of a social is- quantitative methods in data gathering and reporting. sue that traditionally is investigated using qualitative Primary data were gathered through unstructured social science research methods and tools. This paper interviews with six selected elderly men and women. comes from a bigger case study of marginalized el- As operationalized in this study, an elderly is a person derly men and women who are practically living on aged 60 years or older, male or female, of whatever their own in Baguio City, northern Philippines. The marital status, poor2 and practically living alone. study itself is exploratory on two counts; one, it is a Later, some flexibility was allowed to include an first attempt to describe the strategies employed by elderly woman who still cares and provides for her this particular group of elderly in coping with the de- two mentally-challenged dependent adult children. mands of their everyday life in an urban environment and two, which is the focus of this paper, it is an at- The profile of the elderly was described using the tempt to use GIS to integrate and visualize data gener- 2000 Census of Population. Positional data on the ated from key informant interviews. In particular, the participants’ institutional (e.g. hospital, community individual institutional and social sources of support center, marketplace, place of employment) and social as identified by the case study participants were pre- networks (e.g. nearest family member, friends) were sented and analyzed in a GIS environment. Infer- collected using GPS receivers and processed with a ences are drawn particularly on the accessibility of GIS software. Using analysis tools, buffers were sources of support, as well as on the potential hazards introduced around each participant’s dwelling place to (or security) of the participants’ living and working assess the distance between points. A road network conditions. was digitized using a topographic map and a Google Earth image. 1.1 Project site 3. RESULTS AND DISCUSSION The study was carried out in Baguio City, a highly- urbanized, first class city (based on the classification 3.1 Profile of the Elderly in Baguio City of the Department of Interior and Local Government) in northern Philippines. It has an area of 57.49 square The 2000 Census of Population is the most recent kilometres of gently sloping to mountainous terrain. official source of data on the elderly population of It has a relatively cool climate, with temperature that Baguio City. In 2000, the City’s population of can go lower than 10ºC in February and hardly 252,386 was made up of 55.5% in the working age exceeds 26ºC in the summer months of April or May. population and 44.5% dependent population (0-14 and Its population in 2007 was 301,926 and has a 60 and above). Of the dependent population, 4.6% or population density of 44 persons per hectare.1The 11,542 were 60 years and older. The group referred annual population growth rate since 2000 was 2.5% to as “young-old” (adapted from Garfein and Herzog because of a transient student population and a high 1995) or those from 60-69 years old made up 66.43%, rate of migration. Its location as gateway to the the “old-old” (70-79) constituted almost 25% and the northern provinces and proximity to lowland towns “oldest old” (80 and above) were 8.63%. The female makes it an ideal center for education and commerce. elderly in all age groups outnumbered male elderly, It is a major tourist destination and as an urban center, supporting the computed projected life expectancy at has sufficient amenities in communication, health, birth of 73.08 for the Filipino female and 67.83 for social welfare and transportation. the male. In terms of civil status, there were 677 or 5.87% of the total elderly population who were single. These characteristics became the backdrop against A majority was married, and over 31% was widowed. which the situation of the marginalized elderly living A smaller number represented divorced and separated. alone was studied. The setting is also significant Of the single elderly, close to 40% belonged to the considering that in an urban area, relationships tend to 60-64 age bracket; and the remaining 60% was be more impersonal and secondary, compared to a distributed among the other age groups, showing a Philippine rural community wherein a reciprocal co- decrease in frequency count as the age bracket goes existence and caring for one’s relative or neighbour is higher. An insight into living arrangements of the a social responsibility that is commonly observed and elderly can be gleaned from the information on how where care for the young and the elderly is normal and expected. 2 The poverty line or the poverty threshold, which refers to the annual per capita income required or the amount to be spent to 1 Medium Term Development Plan, City of Baguio (2005 – 2010). satisfy basic food requirements and other non-food basic needs, was P16,810 in 2006. This means that the average person in CAR whose annual income falls below this level will be considered poor. Likewise, a family of five would need at least P84,050.00 annually in order not to be considered poor.

Author’s Guidelines – HealthGIS 2009 and church mates, but about to get regular pension from the social security system. Another gets many household members an elderly belongs to. allowance for taking care of pets in the house and Some were found in households with as many as 8 business premises of his guardian, and another stays members; while 4% (474 of the 11,542) were living in his quarters all day and waits for his meals to be alone. Among the elderly living alone, over 50% brought by his daughter. Living arrangements varied were in the 60-69 age group, and the percentage value with one living alone her two-level semi-permanent also tapers as the age group goes higher. house; another has a shack connected to her son’s property, and the other rents a 4-square meter room The distribution of the elderly population by barangay with her two children. (smallest political unit) was generated. Notable is the concentration of elderly citizens in the fifth smallest 3.3 Themes barangay called Imelda R. Marcos, with 2,114; followed by Irisan, the biggest barangay with 1,187. Based on the questions raised by the study, three Among the 20 districts of the city, the highest themes were created: the perception of the elderly concentration of elderly citizens were in District 1 about old age, social support network and map composed of barangays Pucsusan, Gibraltar, Pacdal, generation. Mines View Park and Saint Joseph Village with a total elderly district population of 1,492. There are Perceptions about old age: Aging is an inevitable no known special facilities for the elderly in these process marked by the slowing down of the body’s communities; nor is the terrain especially suitable to systems. The meanings attached by the respondents the physical limitations of the elderly. on old age were initially based on the changes they observed with their bodies. Foremost of these were 3.2 Profile of the Cases the physical changes such as the decline in physical strength as when one has difficulty in lifting objects There were three males and three females involved in and carrying heavy loads, wrinkling of the skin and the study (see Table 1). The youngest was a 60-year poor eyesight. Forgetfulness marked the cognitive old single male, and the oldest was an 86-year old change they experience, while irritability and being once-married female. All the cases were hardly sensitive were the emotional changes they claimed to educated, with two never having gone to school, two now possess. having acquired up to three years in the primary level, and two reaching the secondary level. All of the In relation to the perception of being alone in old age, cases were migrants, one having come from China. the responses manifested reflections about their life experiences. The cases looked at their lives as a Table 1: Demographic Characteristics of the Case showcase of varied situations ranging from happiness, Study Participants contentment, and being free from responsibilities to being burdened. 1. P 60, single, unemployed, some elementary an education, skilled in carpentry, about to Social support network: This refers to the persons receive social security allowance from whom an elderly person seeks assistance in the form of care giving, psychological support and social 2. M 65, never married, Chinese migrant, contact. From the cases, two sources of support were ar never went to school, does chores for a evident. In terms of a social group, the family of rich Chinese family for some allowance, orientation (family where one is born into) and the provided housing and subsistence family of procreation (own family established by a person) both figure in the support network. A major 3. V 86, separated; never went to school; consideration in the exchange relations between the al taught to read by nuns self-employed as elderly and the family members identified is the store keeper, migrant, owns house financial capability of the source. 4. G 70, separated, retired from the mines, Physical proximity between the elderly and the in migrant, stays in a room for free near members of the network is another factor that own children emerged as influencing exchange relations. Other primary social groups that figure in the social support 5. I 71, widow, second year high school, network are the neighbours and the town mates. li peddles vegetables, migrant, rents a house Institutions form another category of social support network. These include government agencies, non- 6. L 84, widow, some elementary education, government organizations and the church. In the case or laundry woman, migrant, owns a shack within son’s lot By way of sources of living, the women were self- employed. The oldest was managing her own retail store and the two others were earning a living on a day to day basis as a vegetable vendor and a laundry woman. The men were basically non-cash earners, with one dependent on the support of family members

Author’s Guidelines – HealthGIS 2009 therefore varied and cannot be said to be habitual or regular. One was visited only when he shouted for of the elderly in this study, these institutions provide help or his meals are brought to him, while others an extension, if not a substitute, for family. communicated with their family members on a daily basis. GIS Application: To complement the qualitative data generated through key informant interviews, the #*VegetableTrading Post ­ social and institutional networks of the case study participants were mapped. The process for generating the maps is shown in Figure 1. Data Collection Google Earth imagery, GPS Points, Topographic maps #*Val's niece #*Mar's family doctor ^_ ^_ ^_#*Ili Topographic Google Earth Image #*#*VaVla'sl/tWowornkmplaat##c*e*eLLoorr's/Fwamorikly #*City m##*a*rMkeatr's workplace 2 Maps GPS points ^_#*S#u*nSshoicniealPW#a*erklGfairne/FOafmficiley Scan, georeference, digitize roads ^_##**PPaann/'Fsacmhuilyrch #*Baguio General Hospital 0 0.3 0.6 1.2 1.8 ^_#*Mar/Family Kilometers Georeferenced map Rectified Figure 2: Summary of institutional and social support image networks of the case study participants Buffer Individual Social and Livelihood. Of the four who were earning a living, (500 m ) Institutional Networks three were working outside the home. One worked within 500 meters to 1 kilometer of her employers, Figure 1: Process to generate the maps one within 1.5 kilometer and another stops in two places at distances of 4 and 8 kilometers. This case in The topographic map of the city and a Google Earth particular either rode a public transport vehicle or image served as base maps to plot the collected GPS walked to reach her places of work. One of the points and to digitize roads. A buffer of 500 meters participants was managing a small retail store at home was introduced around each participant’s dwelling and she would replenish her stock every week by place to estimate distances between points. going to the city market herself. This is 2.5 kilometers away. She is assisted with pricing and Data collected with the GPS receiver were the inventory by a town mate leaving within 500 meters. location of cases, the closest family member or non- member that provide support, places of employment, Health services. All of the cases never went for important places visited regularly, such as the city medical assistance to the local health centers, which market, and institutions, such as the City Social were within up to a kilometer of their dwelling units. Welfare Office and places that provide health Two went for check up and prescription to a member services. Figure 2 presents the data with a buffer of of the surrogate family or a church mate who were 500 meters to the farthest point. within 2 to 3 kilometers of their dwelling units. One visited the public hospital for specialized services, Closest source of support: Four of the cases were such as psychiatric help for her children, while physically living alone. One was living with her two another went to both public and private hospitals if children but they could not be counted as source of hospitalization was needed. support because she was providing for them. One was living within the same house as his niece’s family, but Social welfare. Only one of the six case participants lives independently. Of those physically living alone, has a senior citizen’s identification card from the City three of the cases have access to their married Social Welfare Office. This card enables an elderly children next door, one to her niece within less than a to access discounts from all commercial kilometer and one, to other hired help and to his establishments. The others have not acquired this surrogate family within the same building. Support identification card at all. The social welfare office from these members varied. One was provided meals located in the city was accessible to most of them, but daily while others went to their closest sources of at a distance of 2.5 to 3 kilometers. There were no support when in need of financial support or sick and social welfare employees within the communities require assistance. One of the cases said due to the whom they can consult for credit for livelihood or distance of her niece, she would lock herself at home support for health care. Only one of the six belonged and sleep till well enough to get up again and work. Interaction with source the closes sources of support

Author’s Guidelines – HealthGIS 2009 they have. These will help barangay leaders, for example, to gain a better understanding on the plight to an organization of elderly, but she said she hardly of a special group of clients that they serve and attended their events because of her work. eventually influence their public policy to favour the elderly who live alone. Recreation and religious affiliation. Of the six, only one appeared to regularly go to the park for For services to be systematic and to reach the most recreation. The nearest park was 2 kilometers away. vulnerable groups, a necessary first step, however is He was also the only one with a strong support from a to undertake a survey to locate the elderly and profile local church, which was within 500 meters of his them. The use of GIS becomes all the more useful, home. especially if eventually the elderly will be tracked, monitored and properly assisted. The generally hilly terrain of the city was another factor that was considered. Five of the elderly were ACKNOWLEDGEMENT living in hilly areas where careful navigation, even to an agile young person was required. This paper came from an on-going study entitled, “Growing Old Alone: An Exploratory Study of 4. CONCLUSION AND RECOMMENDATIONS Childless Elderly Men and Women in Baguio City.” It is funded by the Cordillera Studies Center, UP There is a lesson that can be gleaned from the study Baguio under its Research Fellowship Program. The of the experiences of the elderly men and women who co-researchers of this project, namely Prof. Ma. participated in this research: the support that they Cecilia San Luis and Prof. Ma. Ana Diaz from the need in this stage of their lives cannot be reduced to College of Social Sciences in the same institution only one type. It is not enough that they be provided provided the disciplinal expertise in Sociology and meals, but an affirmation of the self and a sustained Psychology. social interaction between the elderly and the social support network are as important. In the absence of REFERENCES children who could provide different forms of support, the significant others identified were other Garfein, Adam J. and Herzog, A. Regula, 1995, Ro- family members and town mates/neighbours. The bust Aging among the Young-Old, Old-Old, and study showed that marginalized elderly live a difficult Oldest-Old. Journal of Gerontology: SOCIAL life, more so those who live alone. Their access to SCIENCES, Vol. 50B, No. 2, S77-S87. support, both social and institutional, is not within easy reach. For most of the cases, the level of effort Guan, Lee Hock, 2008, Introduction. In Ageing in to get help is much more than if they have family who Southeast and East Asia, edited by Lee Hock are living with them. Guan (Singapore: Institute of Southeast Asian Studies). The maps that were generated in the study complemented these findings. The maps magnified National Statistics Office, 2000, Demographic and the findings that for some, the immediate social Housing Characteristics, Census of Population and sources of support are relatively inaccessible. They Housing, Baguio City. Report No. 2-11 N, likewise highlighted the physical distance of Volume I. institutions that could provide alternate help for them in the absence of a social support network. Under Natividad, J.N., 2008, Family and housing conditions these circumstances, the elderly is very vulnerable. of the elderly in Southeast Asia: Living arrange- ments as social support. In Ageing in Southeast This paper presented only one aspect of analyzing the and East Asia, edited by Lee Hock Guan (Singa- social and institutional structures in which the elderly pore: Institute of Southeast Asian Studies). lived. More analysis can be done by incorporating other variables like slope, other institutions, Tinker, A., 1981. The elderly in modern society, transportation networks and the like. (New York: Longman Group Limited). A study of elderly men and women wanting of social Tornstam, L. 2003. Gerotranscendence from young support can provide a strong grounding for a more old age to old old age. Online publication from useful definition of care-giving and proper planning The Social Gerontology Group, Uppsala. URL: for the elderly. The visual representation of data is an http://www.soc.uu.se/publications/fulltext/gtran- effective way to convey information in a non- soldold.pdf. technical manner to a variety of policy stakeholders. Maps can help them understand how the elderly cope; http://nscb.gov.ph/rucar/ the levels of risk (or security) that they are in, the amount of time devoted to activities that they do in their daily lives and the limited social networks that

Third International Conference on HealthGIS 2009 July 24-26, 2009, Hyderabad, India THE RELATIONSHIP BETWEEN INVERSION AND TEHRAN AIR POLLUTANTS BY USING MULTIPLE VARIATION REGRESSION TEST (SYNCHRONOUS ENTRANCE METHOD) Saeid kamyabi * and Mahnaz Parvazi * Department Geography, Islamic Azad University of Semnan Branch, IRAN Department Geography, Islamic Azad University of shahr-e-ray Branch, IRAN [email protected], Mahnaz_ Parvazi @yahoo.com ABSTRACT: Air pollution is one of today’s modern life phenomenon's and the most complicated and critical problems of man’s societies, specially in big cities and urban areas of the world. Tehran is one of these urban areas which is facing this problem as a result of several situations and factors. Because of the topography and the natural factors in the north and the south and also major winds in the east and the west, temperature inversion during cold period, low falls which can not clean the air, growth of the population on the other hand and the development of industrial activities using million litters of fuels in various resources, increasing number of vehicles and house fuels as well, It has become on of the most polluted cities of the world. The result of the above mentioned problems are heart-failor,respirator diseases, lung problems,nurve illnesses, chronic headaches and many other diseases. In this survey, the relationship between air pollutants and inversion and its impact on mortality due to heart-failures in Tehran and in the ten-year statistic period from 1997-2007was reviewed in a daily manner. The statistic related to mortality numbers due to heart-failure and lung diseases, station of Vila, Azadi, Bahman, Pardisan, Tajrish, Gholhak and the Sorkhe Hesar for different days of a year were obtained from the related organizations.They were processed by Excel,GIS and Spss softwares and the relationship between heart-failure and inversion and air pollutants was analyzed using statistic methods. The results considerably show that heart failure is meaningfully related to the growth of air pollution like Carbon monoxide(CO), suspended particles(PM), Dust, Nitro oxide (NO), Ozone (O3) and the inversion day during a ten-year period diseases with Co and NO./001level,PM and O3./05 and lung diseases with NO,PM, and Dust ./05 these present in during 1997-2007statistical period. KEY WORDS: Air pollution, Urban, Climate, Inversion, lung failures, GIS, Tehran 1. INTRODUCTION World Health Organization (WHO) researches have shown that 1.4 billion people of city dwellers of the Air pollution as the result of combustion sources is world breathe an air whose pollution is beyond the accompanied with a broad range of acute and chronic standard level of WHO and is not healthy. Automobiles health troubles which change depending on the type of are considered as one of the most important sources of pollutant. International health report indicates that air air pollution in Europe whose effects on citizens’ health pollutant particles (particles which are so small that may are rapidly increasing (Pourianejad, 2005, P 88). easily enter the lung) are the important respiratory factor and have an inseparable relation with other cases of Tehran is among those capitals which have different death rate due to cardiac and pulmonary troubles administrative, political, commercial, industrial, cultural (Pourianejad, 2005, P. 87). and artistic functions. Rapid increase of population, rural immigrations, development of factories, The observed effects of air pollution on health include accumulation of vehicles (Mohammadi, 2002, P. 154) increase of respiratory signs, decrease of lung activity, and high consumption of fossil fuels as well as low increased number of individuals bedridden in the quality of consuming fuels in transportation sector, lack hospitals due to respiratory, cardiac and vascular of any clean and unpolluted‫و‬d transportation system, diseases, increase of absence from work and school due economical and social weakness, unsubstantial to respiratory disease, increase of death rate of cardiac management in transportation sector (Conference of Air and respiratory patients (Pop et. al, 1995, PP 427-480) Pollution and Its Effects on Health, 2005, P. 19), and mild anemia of unspecific activat6ion of safety topographical conditions and natural factors have system and delay in completion of bones in children changed Tehran into one of the most polluted cities of (Whichman et. al, 1995, P 29). the world.

Since air pollution has the most effects on organs such 6,700,000 (table 1) and its population growth has as heart and lung, the effect of Tehran air pollutants on decreased from 6.05% in 1956 to 1.3% in 1996 (table 2). pulmonary diseases during statistical years of 1997-2007 Decrease of population growth rate in Tehran doesn’t has been investigated in this paper. mean the decrease of Tehran population, rather it may be considered as the result of population movement Figure1Stations for measurement Air Pollution of study from central parts of the city to its suburbs (Kargar, area (Tehran, Iran) 2004, P. 113). 2 RESEARCH GOALS As a result, it can be said that the increase and high rate of population growth as well as rush of immigrants The goal of this research is to investigate the relation toward large cities including Tehran has caused the large between air pollution and inversion days and their cities of the country and centres of provinces to face effects on the deceased due to cardiac and pulmonary phenomena such as living in suburbs, formation of p diseases during the 10-year statistical period of 1996 to henomena such as living in suburbs, formation of poor- 2006. settler localities, air pollution, traffic and other The role of geographical factors in Tehran air pollution: environmental problems besides rapid formation of new Geographical factors are divided into two groups as centres of cities (Malekoutian, 1978, P. 18). follows: Pollutant sources of Tehran air: Pollutant sources of Tehran air are divided into two 1. Natural geography factors general groups: - Situation, extension and physical extension of 1. Natural sources Tehran 2. Human sources - Topographical situation of Tehran 1. Natural sources are those sources that man does not - Altitude from the sea interfere in their generation, such as dust due to wind, - Tehran climate (rainfall, wind, distribution of ashes and gases of ozone volcano due to thunder and lightning, smoke, gas and ashes due to conflagration of temperature regime distribution, relevant hu- forests, natural radioactivity such as radon gas from the midity distribution, temperature reversal, pres- earth surface (Atabi, 2003, P. 3). sure patterns). 2. Human sources are those sources that man interferes 2. Human geography factors in their generation, such as motor vehicles, industrial - Population increase and development of the units and power plants, heating, home and commercial city sources, miscellaneous sources such as garbage burning, Natural geography factors may be effective in the tire burning, bitumen heating and propagation of intensification of air pollution of the city but it is human different kinds of chemical products (Moharamnejad, factors which cause air pollution of the city whose 1984, P. 25). greatest and most important factor is population and its accumulation in a definite area. 3. DATA PROCESS AND ANALYSIS In the process of its spatial / skeleton development, Tehran has become 25 times broader than 1891 to 1996, Analyses related to the investigation of relation i.e., during a 105-year period, and its population has between the condition of inversion variables and cardiac increased 33 times compared to the beginning of period. and pulmonary diseases and PM and CO of pollutants. At the present time, the boundary of Greater Tehran has In this section, the analyses related to the investigation an area of 1,000 km2 which is extended from the farthest between the conditions of inversion variables, cardiac end of north to the south maximum 40 km2 and from the and pulmonary diseases and pollutants through trilateral farthest end of east to the west maximum 70 km2 variance analysis are dealt with. Considering table 1, it (Kargar, 2004, P. 146). can be said that cardiac disease indicates a significant relation (in a level of a = 0.001) with CO contamination, One of the important factors of spatial development of in the manner that in case of CO contamination, the rate Tehran is population growth which is the result of two of cardiac death (with an average of 7.34) will be more factors, natural growth and increase of immigration. than non-death rate (with an average of 5.74) (figure 1). During the statistical period of 1956 to 1996, Tehran population has increased from 1,600,000 individuals to Considering figure2, it can be said that the rate of heart death has a higher peak compared to non-death rate and therefore it can be said that cardiac death rate is higher. Considering figure 3, it can be said that pulmonary death rate in the days of cardiac deaths has a higher peak compared to pulmonary non-death. Therefore, it can be said that in case of any PM pollutant, pulmonary and cardiac death rates are higher. Considering table 2, it can be said that all the main effects (inversion, cardiac

and pulmonary diseases) indicate a significant relation their improvement up to the international (in a level of a = 0.05) with PM pollutant, in the manner standard. that in case of any PM pollutant, inversion 3. Improvement of passageways network, devel- intensification (with an average of 112.51) is higher than opment of passageways, making multi storey when the inversion intensification is low (with an and round-the-clock car parks in different average of 91.81). parts of the city. Also in case of any PM pollutant, death rate due to 4. Development of clean public transportation cardiac disease (with an average of 102.49) will be fleet and completion of 7 lines of Tehran higher than non-death (with an average of 90.50). metro in the shortest possible time as well as Ultimately, in case of any PM pollutant, non-death rate supply of relevant sources. (with an average of 101.64) will be higher than death 5. Regular and periodical technical inspection of rate due to pulmonary disease (with an average of cars, decisive approach with the smoky cars 97.76), (figures 2 & 3). Considering figure 3, it can be which don’t have any technical inspection re- said that pulmonary death rate in the days of no cardiac port and reinforcement of traffic police. death has a higher peak compared to pulmonary non- 6. To provide facilities for reducing unnecessary death rate. Therefore it can be said that whenever there urban trips (development of different services is a PM pollutant, pulmonary death rate is higher. such as post, telecommunication, etc. 7. Improvement of public culture regarding air 4. CONCLUSION pollution issues and making people familiar with their duties for reducing pollutants. To investigate the relation beween inversion and 8. Control of the pollution resulted from con- pollution in spss software, trilateral variance analysis sumption of fossil fuels as well as control of model was used in the manner that inversion and cardiac pollution arising from different processes in and pulmonary deaths were considered as independent industry (use of tools and systems and filters or predictor variable and the pollutants were identified which are called Electro filters). as dependent or criterion variable. Finally, the relation 9. For the executive measures, population attrac- of variables was dealt with. The obtained results tion centres such as industries, political and between inversion, pollutants and death rate are as economical poles must be distributed through- follows: out the country instead of concentration in Tehran. Prevention from the development of 1. There is a significant relation in a level of activities of institutes and ministries in Tehran 0.001 between inversion and CO pollutant and and providing necessary groundwork's for cardiac death, in other words, there is the pre- their development throughout the country diction capability of CO pollutant through in- version. In case of CO pollutant, cardiac death ACKNOWLEDGEMENTS rate is higher but positive significant relation with pulmonary death is weak. Acknowledgements of support for vice research Islamic Azad University of Semnan Branch, IRAN.. 2. There is a significant relation in a level of = 0.001 between inversion and PM pollutant, in other words, there is the prediction capability of PM pollutant through inversion. In case of PM pollutant, cardiac and pulmonary death rates are higher. There is a positive significant rela- tion with cardiac and pulmonary deaths but the relation with cardiac death is higher. 5. SUGGESTIONS AND SOLLUTIONS FOR PREVENTION FROM AIR POLLUTION 1. Tehran is a city with a potential resistance, es- pecially during cold months of the year. Its purifying rainfalls are low and have not been distributed in all months of the year. The sweeper winds also don’t blow anywhere and everyday. Therefore, polluted activities are limited to the city and its suburbs, especially west and south parts. 2. Correction and standardization of new auto- mobiles and motorcycles based on approved environmental standards of the country and

Table 1 the relation of inversion variables and cardiac and ulmonary diseapses with CO pollutant through trilateral variance analysis Effects Source of Total Freed Averag F Level Level Main Changes Squar om e of 0.59 of Effects Inversion es Degre Square Signifi Simple Cardiac Disease 65.66 e s cance Interaction Pulmonary Disease 1 0.44 Complicated Interaction of Inversion with 65.66 Interaction cardiac disease Interaction of Inversion with 2449.8 1 2449.86 22.25 *** pulmonary disease 6 0.001 Interaction of cardiac disease with 1 5.20 0.04 pulmonary disease 5.20 0.82 Interaction of inversion, 0 00 - - cardiac and pulmonary diseases 18.06 1 18.06 0.16 0.68 0 - 0 00 - - 00 - Table 2 the relation of inversion, cardiac and pulmonary diseases with PM pollutants through trilateral variance analysis Source of Total Freedom Average of F Level Level of Changes Effects Squares Degree Squares Significance Main Inversion 37638.71 1 37638.71 17.22 *** Effects Cardiac Disease 10680.95 1 10680.95 4.88 0.001 8367.60 3.82 Pulmonary Disease 8367.60 1 0 0 * 0.02 Interaction of 0 0 Inversion with cardiac * disease 0.02 1 0.05 0 0 Interaction 0 0 - of Inversion with pulmonary disease Simple 110.00 0.02 0.01 - Interactio Interaction of cardiac disease 105.00 0 0 0.99 n with pulmonary disease 100.00 0 0- Complicat Interaction 95.00 ed of inversion, cardiac and 90.00 Interactio pulmonary diseases 85.00 n Estimated Marginal Means of co Estimated Marginal Means of pm at diseas.r.r = fot 7.50 7.20 Estimated Marginal Means Estimated Marginal Means6.90 diseas.g.r fot 6.60 nonfot 6.30 6.00 5.70 fot nonfot inversion noninversion inversion Non- estimable means are not plotted Figure 2. Final estimation of averages from CO Figure 3. Final estimation of averages from PM pollutants pollutants REFERENCES

Afshar, Mojgan, Investigation and recognition of Tehran Heating Island, Master’s degree thesis, Islamic Azad University, Central Tehran branch, 1999. Atabi, Farideh, Air pollutants dispersion modelling by GIS tool for using in environmental assessment, PhD thesis, Islamic Azad University, sciences and researches department, 2003. Kargar, Bahman, Urban safety (an assessment of the ef- ficiency of disciplinary and security forces in urban management system), Armed forces geographical organ- ization publications, 2004. Moharamnejad, Naser, An introduction to air pollution, PhD thesis, Islamic Azad University, sciences and re- searches department, 1984. Malekoutian, Mohammad, Air pollution, Iran Azad Uni- versity, 1978. Mohammadi, Hosseinmorad, The effect of climatic ele- ments and Tehran air pollutants on asthma disease dur- ing 1999-1995, scientific research magazine of faculty of literature and human sciences of Tehran University, 2002. Nourieh, Nafiseh, Investigation of MTBE effect existing in the air of cities on people’s health and method of its measurement and assessment, Conference of air pollu- tion and its effects in health, Zist Mohit Pak Institute, 2005. Pourianejad, Fatemeh, The relation of climatic elements and air pollutants of Tehran with the deceased due to bronchitis disease, Master’s degree thesis, Islamic Azad University, sciences and researches department, 2005. Air Quality Control Co., information of clean air, 2003. Rahmatizadeh,shima,other,The use data base in air quality management,Geomatic83 Safavi, Seyed Rahim, An introduction to martial geo- graphy of Iran, Armed forces geographical organization publications, 2002.



FILARIASIS MONITORING AND VISUALIZATION SYSTEM P. Sampath Kumar1, Y.V.S. Murthy2, M. Srinivasa Rao3, K. Sri Ram3, K. Madhususudhan Rao3 and U. S. N. Murty3 1, 2 Geoinformatics Division, National Remote Sensing Centre, Hyderabad, India 3Biology Division, Indian Institute of Chemical Technology, Hyderabad, India   ABSTRACT: A problem can have innumerable factors with various impact levels. Identifying the factors and their relationship can give closer picture of the problem. Public health is one of perennial complex problem for humanity with increasing factors. The problems of health are increasing in both spatial and temporal dimension due to increased risk of disease transmission fuelled by developmental activities, demographic changes and hectic life style. So, prediction of the cause and effect is becoming extremely complex. However, with the advancement of the technologies like electronics, computers, satellites etc, the scope of data collection, analysis, tracking and monitoring is improved. Satellite Remote Sensing (RS), Global Positioning System (GPS) and Geographical Information System (GIS) are the amalgamation of various technologies opening avenues to assist the public health officials in locating, visualizing, planning, and monitoring. Bancroftian filariasis was considered as a rural problem during 1962-71. National Filarial Control Program (NFCP) covers only one tenth of the population by its control operations. Filariasis is a non-fatal vector-borne disease which comprises a dynamic interaction between the pathogen, the vector, the vertebrate host and the environment. To control the disease it is necessary to have a complete database which consists of Entomological, Epidemiological and Socio-economic factors. Based on various prospective community-based rural bancroftian endemic studies, the study areas are selected. These study areas also have very high vector proliferation due to lack of proper communication between the community and the Public Health Centers. Extensive ground surveys are conducted to collect the spatial, Epidemiological, Entomological and Socio-economic data. These data sets are analysed to obtain four parameters viz., Infection rate, infectivity rate, Per Man Hour Density (PMHD) and Micro Filarial (MF) Rate. Further, these parameters are attached with administrative boundaries in spatial domain for their visualization in GIS environment. The application developed helps not only monitoring Filariasis, but also in control operations. KEY WORDS: GIS, GPS, Filarisis, Customization, Visualization 1. INTRODUCTION 2. LITERATURE REVIEW Location mapping and spatial analysis are vital in various Lymphatic filariasis (LF) is disabling parasitic disease domains particularly public health which is influenced by that has been identified by the World Health Organization socio economic and environmental factors. The studies (WHO) as a major public health problem with an are carried out in general and through classical method of increasing prevalence worldwide (WHO 1992). At approaches for smaller extent. Due to the rapid present, about 20% of the world's population is at risk of development of various technologies and globalization, infection, over one third of the population at risk lives on the propagation of an event is very quick to wider regions the Indian sub-continent (Michael et al., 1996) and in on surface of earth including the diseases. To meet the India about 553 million people across 22 States/Union challenges due to rapid interdisciplinary solutions are Territories are known to be endemic (Reddy et al., 2000). becoming essential. Spatial technologies are resultant of amalgamation of various advanced engineering and Although LF does not cause immediate mortality, the science domains primarily consisting of electronics, associated severe morbidity has resulted in it being computer science, civil engineering and earth sciences recognized as the second leading cause of disability which have aided for the advancement of Satellite worldwide (WHO 1995). The two most common chronic Remote Sensing, Geographical Information System and manifestations of the disease hydrocoele and Global Positioning System. Increased awareness of lymphoedema cause socio-psychological problems to recent advances in GIS, mapping has created new patients and their families (Ramaiah et al 1999). Chronic opportunities for public health administrators to enhance disease is debilitating, leading to a restriction in the their planning, analysis and monitoring capabilities. duration and capacity to work and to changes in activity These health services include health care services, patterns (Evans et al., 1993, Ramaiah et al., 2000). administrative services, monitoring for various health programs, identify environmental, socio-economic Since the health issues are always related to space and geographic and demographic factors and other risk time, and therefore it is ideal to link Geographical factors etc, (B. Suresh, 2002). Information Systems (GIS) with Health Information. GIS provide ideal platforms for the convergence of disease- specific information and their analyses in relation to

population settlements, surrounding social and health refers to the scientific study of insects (wikipedia). This services and the natural environment. They are highly survey was carried out with the help of a pre-defined suitable for analysing epidemiological data, revealing survey format, mosquito trapping and Garmin GPS 12 trends and interrelationships that would be more difficult instruments. to discover in tabular format. Moreover GIS allows policy makers to easily visualize problems in relation to The parameters that were surveyed include primarily the existing health and social services and the natural collection of blood samples from individuals and the environment and so more effectively target resources. catching the mosquitoes in the vicinity of ground surveys. While collecting the blood samples individual persons 3. OBJECTIVES details such as Name, age, sex, economical status etc Based on the requirement projected by IICT, a GIS based were also noted for the necessary analysis. In addition the monitoring and visualization system called Filariasis type of the house and meteorological data like Monitoring and Visualization System (FMVS) for vector temperature, rainfall, humidity, wind speed in the borne diseases planned and developed named. The selected villages were collected. Apart from this GPS overall objective of FMVS is to store, organize and instruments were used to locate and identify the villages analyze the filariasis data in spatial environment to view, geographically as shown in Figure 2. query and visualize the parameters associated with the disease. Figure 2. GPS raw data collected from Field 4. STUDY AREA 5.2 Data Analysis This disabling disease is prevalent in urban and rural areas affecting people of all ages, both sexes, particularly The field survey data was brought into a digital database those of low socioeconomic status (Crompton et al., for further analysis. The blood samples and the 2003), and living in the areas with improper drainage mosquitoes were analyzed in the lab to obtain their system. Some of the earlier studies carried out in East infective nature. Further, the data was analyzed used data Godavari district found the disease acute and chronic in mining and classification tools such as Classification and the region. Based on these studies three additional Regression Tree (CART) and Self Organizing Maps districts are selected which are geographically distributed (SOM) which are commercially available. in the state of Andhra Pradesh namely West Godavari, Karimnagar and Chittoor. About 120 villages are selected CART is used to analyze either discrete (classification) or across these four districts for the focused study. The continuous (regression) data sets. The key elements of spatial location of these districts is given in Figure 1. CART analysis are set of rules for splitting each node in a tree; deciding when a tree is complete; and assigning each Figure 1. Location Map showing the study area districts terminal node to a class outcome (or predicted value for regression). CART 5.0 could rank the predictor variables 5. DATA PREPERATION according to their level of influence on the target variable 5.1 Data Collection (Steinberg, D., and P. Colla. 1995). Thus, “Age”, “Habitat”, “Sex” and “House Type” were found to be The data pertaining to epidemiology, entomology and influencing the target variable in the descending order. spatial locations are collected in the intensive ground Self-organizing maps (SOM) is a data clustering survey conducted in villages. Epidemiology refers to the technique, which reduces the dimensions of data through study of factors affecting the health whereas entomology the use of self-organizing neural networks (Kohonen, 1995). Through these analyses four parameters are obtained per- man hour density (PMHD), Microfilaria (MF) rate,

Infection rate and Infectivity rate and are represented in three intensity levels as mentioned in Table 1. Table 1. Intensity classification for parameters Parameter Low Medium High PMHD 0-10 10-40 >40 0-5 5.1-10 >10 Infection rate 0-0.2 0.2-1 >1 Infectivity rate 1-5 5-10 >10 MF rate The spatial data used for developing this application is Figure 4. Application architecture the village administrative boundaries to which above four parameters are attached. For matching of gourd survey 6.3 Software Modules village locations collected from GPS with the spatial boundaries from authenticated sources were carried out To meet the objectives according to the designed with the help of available GPS locations (Figure 3). architecture the application is dividend into three basic modules namely Data access module, Visualization Figure 3. Village Boundary Identification module and thematic representation module. Once the village boundaries identified then they are The data access module fetches the data stored in ESRI linked based on the name as common field between Personal Geodatabase (GDB) which contains spatial spatial and filariasis non-spatial data bases in a specific layers such as Village boundaries. The data access format suitable for the application development. This module invokes the map window(s) for the selected form of data was ingested into the application for district data user can choose the there by data through the visualization and query purposes. combo box provided in the interface. The visualization Module loads and displays the selected data in the map 6. APPLICATION DEVELOPMENT window(s) for viewing and navigation at various levels. 6.1 Tools and Technologies Functionalities such as panning, zooming, full extent etc provided in the tool bar. A simple data query tool is also Customization of GIS application is the process of made available to view the attributes of a selected village. leveraging the available functionalities in a desired In Thematic representation module four pre-defined manner from software development kits (SDKs). ArcGIS parameters are used to represent their intensities. Unique Engine is one such kit of GIS functionalities which can value rendering functionality is used to represent them in be accessed through any high level programming different colours by reading the corresponding attribute of languages such as C++, VB, .Net, Java etc. For the each village. The village without data will have no current application .net is used. colour. 6.2 Architecture The unique feature of this application is to display data in four geometrically synchronized map windows. Each To incorporate functionalities such as data access, data parameter is displayed in different window for the same visualization and thematic representation based on study area. A location map is provided to show the study attributes, Description of various modules is given below area in the state map. and the architecture of the application is depicted in the Figure 4. These modules are accessed through a user friendly Graphic User Interface (GUI), consisting of splash screen (Figure 5) and Main window (Figure 6). The main window is divided into four map display windows, one location map window, a navigation tool bar and a layer list showing their corresponding legends.

Figure 5. Splash screen Michael E, Bundy DAP and Grenfell BT 1996, Re- assessing the global prevalence and distribution of Figure 6. Main window displaying Chittoor district lymphatic filariasis. Parasitology, 112:409-428. 7. RESULTS & CONCLUSIONS Onapa, A.W., Simonsen, P.E., Pedersen, E.M., Okello, Customized GIS solutions are very useful in the health D.O., 2001. Lymphatic filariasis in Uganda: baseline domain for improving the data management, visualisation investigations in Lira, Soroti and Katakwi districts. and decision making. It is achieved by the visualisation of Trans. Roy. Soc. Trop. Med. Hyg. 95, 161–167. spatial distribution of diseases, facilitates, monitoring and appraisal of the effectiveness of health programmes. Ramaiah KD, Radhamani MP, John KR, Evans DB, There is scope further for spatial analysis, queries and Guyatt H, Joseph A, Datta M, Vanamail P, 2000 : generating reports. Scenarios can also be developed based The impact of lymphatic filariasis on labour inputs in on the analytical results of the factors that affect disease southern India: results of a multi-site study. Annals for the villages with out ground survey using the socio- of Tropical -Medicine and Parasitology, 94:353-364. economic, Land Use Land Cover data etc. Acknowledgements: Reddy GS, Venkatesvarlu N, Das PK, Vanamail P, The authors thank Dr. P. S. Roy Deputy Director Vijayan SK, Pani SP. 2000, Tolerability and efficacy (RS&GIS-AA) and Dr. V. Raghavaswamy, GD, USGIG, of single dose diethylcarbamazine (DEC) or National Remote Sensing Centre (ISRO), Department of ivenmectin the clearance of Wuchereria bancrofti Space, Government of India, for their encouragement microfilaraemia at Pondicherry, south India. Trop. while carrying out this work. Med. Int. Health; 5: 779-85. REFERENCES Reuben R, Tewari SC, Hiriyan J, Akiyama J. (1994) Illustrated keys to species of Culex (CULEX) Evans DB, Gelband H, Vlassoff C, 1993 : Social and associated with Japanese Encephalitis in South-East economic factors and the control of lymphatic Asia (Diptera: CULICIDAE). Mos. Sys, 26 (2): 75 – filariasis: a review. Acta Tropica, 53:1-26. 96. Suresh B. 2002, GIS as a tool for monitoring Health Management Information System, Proceedings Map India 2002. Yasuno, M., Rajagopalan, P.K., Kazmi, S.J & La Brecque, G.C, 1977, Seasonal changes in larval habitats and population density of Culex fatigans in Delhi villages. Ind. J. Med. Res, 65 (Suppl.): 52-64. Sasa M. 1976 Human Filariasis: a global survey of epidemiology and control. Tokyo: University of Tokyo press, 663-734. Kohonen, 2001, Self Organizing Maps, Springer Series in Information Sciences, Vol. 30, Springer, Berlin, Heidelberg, New York, 1995, 1997, 2001. Biswas G, Raina VK, Rao CK, 1996 : Revised strategy for the control of lymphatic filariasis in India, New Delhi. National Institute of Communicable diseases and National Malaria Eradication program 43 Crompton DWT, Montresor A, Nesheim NC, Savioli L, 2003 : Controlling disease due to helminth infections. World Health Organization, Geneva;. Detinova, T.S., 1962. Age-grading methods in Diptera of medical importance. In:World Health Organization Monograph Series No. 47. World Health Organization, Geneva.

Project report 2008: An integrated information system for the control of bancroftian filariasis in Andhra Pradesh, Submitted to Ministry of Communication & Information Technology Govt. of India, New Delhi WHO, 1987. Control of lymphatic filariasis. In: A Manual for Health Personnel. World Health Organization, Geneva. WHO 1992, Lymphatic filariasis: the disease and its control. World Health Organization Geneva: Technical Report Series, 821:1-71. WHO 1995, Bridging the Gap. In The World Health Report, Geneva. Steinberg, D., and P. Colla. 1995. CART: Tree-structured non-parametric data analysis. San Diego, Calif., U.S.A.: Salford Systems. http://www.salfordsystems.com

USING GIS IN EXPLANTING SPATIAL DISTRIBUTION OF BRUCELLOSIS IN AN ENDEMIC DISTRICT IN IRAN Ali-Akbar Haghdoost1, Leo Kawaguchi2, Ali Mirzazadeh1*, Hossein Rashidi3, Afshin Sarafinejad1, Ali-Reza Baniasadi4, Clive Davice2. 1 Kerman Medical University of Medical Sciences, 2 London School of Hygiene and Tropical Medicine, 3 Provincial Vet office in Kerman Province, 4 Management and Planning Organization in Kerman Province, Physiology Research Centre, University of Medical Sciences and Health Services, Kerman, Iran. P.O.Box:76175-584 Email: [email protected] KEY WORDS: Risk map, GIS, Human brucellosis, livestock 1. INTRODUCTION under supervision of different ministries, and there is no direct link between these datasets. However, we believe There is a well established primary health care (PHC) that health system can response much faster and on time delivery system in Iran, particularly in rural area. More to outbreaks if these data can be shared effectively.The than 16000 health houses in main villages around the spatial modelling using the capacity of geographical country cover close to 95% of the rural population[1, 2]. information system (GIS) is directly applicable to Health workers not only deliver primary health cares, but understand the spatial variation of disease such as also keep health records of people in their catchments brucellosis, and its relationship to environmental factors areas and play an important role in surveillance system. (livestock's parameters). Although, GIS is a well known In spite of well-established PHC, Brucellosis still is an tool in health system in developed countries[4], its' endemic disease in Iran, particularly in rural areas. application is very limited in developing courtiers such as Unfortunately, Iran is one of the top five countries with Iran. In this study, we linked the brucellosis surveillance high incidence of brucellosis. Although the current data of human and animals in an endemic district in surveillance system in Iran misses a considerable central part of Iran, Bardsir, using GIS methods. We proportion of cases, more than 17,000 cases were aimed to find determinates of human brucellosis and reported in 2003, second only to Syria [3]. check the feasibility of such a model in generation, by linking between human and animal data. Despite of the bilateral effects of human and animal 2. MATERIALS AND METHOD diseases, unfortunately it is not easy to link these datasets in Iran. Well known diseases such as TB, Brucellosis, Rabies, Hydatic cysts, and even new treats such as Bird flu are examples of common diseases between human and animals. The human and animal datasets are compiling

Study area location of epidemiological units. There were substantial mismatches in the spatial distribution of epidemiological Bardsir is located in Kerman Province, south-eastern units and villages, since animals were mostly kept outside Iran, with 10,000 square kilometres area. It's population the villages. Therefore, we used kernel smoothing (an is about 50,000 and around half of them live in rural interpolation technique) to find the best estimate of areas. It consists of steppes and semiarid plateaus, animal density in each village using all available covered by brown soil that supports grassy vegetation. information of animals around villages. Using sensitivity The height ranges from 1900 to 2900 m. Main industries analysis, a bandwidth of 3,000 meters was adopted as the in Bardsir rural areas are agriculture and husbandry. most appropriate bandwidth in the kernel smoothing. Wheat, corn, sunflowers, pistachios and fruits are cultivated[5]. According to the veterinary census data, The villages were classified as endemic if at least one about 25,000 cattle and 500,000 sheep / goats are raised case of human brucellosis was reported between 2002 in the study area. Bardsir district is one of the endemic and 2004. regions in Iran for brucellosis. Health facilities in Bardsir are mostly provided by A village with easy access to health facilities was defined governmental system. Health workers in rural health as a village with at least one of these facilities: Health houses survey people actively and record diseases such as House, midwives, nurses, pharmacy and trained brucellosis and report suspected/diagnosed human cases veterinary staff. The crude and adjusted odds ratios were monthly to the centre of province. In parallel, district calculated to estimate the determinant factors of endemic veterinary office surveys livestock's populations and their villages. The final model was selected by backward related conditions such as diseases and pastures. The stepwise method using logistic regression, by excluding number of livestock is reported for each ‘epidemiological least significant determinant factors one by one. We used unit’, which is defined as an interacting herd of livestock. the natural logarithm of human and livestock population, However, the numbers of inoculated vaccines has not entered in the models as explanatory variables because of being recorded classified accurately to epidemiological their positive skewness; we add one to the populations of units. The veterinary data is being reported annually. livestock to deal with zeros. All geographical analyses were done by the ArcView GIS software ver3.3 and its Datasets and electronic maps extension package ArcView Spatial Analyst (Environmental System Research Institute, Inc.). The Brucellosis cases in 2002-4 have been collected from statistical analysis was done in stata version 8. district Health offices, classified by village. Human demographic information of villages included number of 3. RESULTS households, population in each village, number of literate people, and data on employment were estimated from During 2002 and 2004, 97 patients were reported in 34 1996-1997 census data [5]. Availability of Health House, villages; that means, 10.9% of villages were endemic (34 midwives, nurses, pharmacy and trained veterinary staff out of 311 villages) (Table 1). The annual incidence of were also counted as health facilities. The annual census human brucellosis was 141.6 per 100,000 inhabitants data of sheep, goats, cattle and other livestock is collected (95% CI: 134.4-149.2). from district veterinary office. The number of livestock has been reported to each ‘epidemiological unit’. The clusters of endemic villages and high density spots Electronic map of the district was received form the for cattle, not for sheep were spatial matched. Most of National Statistics Office in Tehran in UTM format in endemic villages were located in the north and south of scale of 1:50,000. Provincial veterinary office also the district. Also, three main clusters with the highest provided coordinates of epidemiological units. cattle density were in north and one in south. There were much more clusters with high density of sheep, mostly Calculating and Statistical methods Using add event function in ArcView software, a new layer was added to the electronic maps to illustrate the

located in north and south; however, they were distributed within wider the district than the cluster of cattle (figure 1). Human and cattle populations were significant Figure 1 Animal population density map in Bardsir determinant factors for being endemic villages. The A: sheep, B: cattle. proportion increased with density of cattle, and reached up to 29% in villages where the density was higher than Both maps were created by setting bandwidth of 3,000 20 cattle per square kilometres. Even after adjustment for metres. Location of village with brucellosis (red) and other potential risk factors (Socio-economic indicators, Education and Geographic factors), human and cattle without disease (blue) is overlaid on the maps. population were positive determinant factors for being endemic villages (human population: OR=1.94, cattle 4. DISCUSSIONS population: OR=1.81) (Table 2). Our study showed that we could generate informative risk On the other hand, the density of sheep did not show a maps of brucellosis using health and veteran data which linear relationship with occurrence of human disease in a may improve the quality of control programme in Iran. village (Table 1). Having adjusted, a negative association Our findings imply that Human Brucellosis is highly was observed between population of sheep and risk of associated with population size and distribution of cattle endemicity and it has borderline significancy. (OR=0.75, in rural areas of Bardsir district, while we could not find 95% CI: 0.56-1.01). (Table 2). any associations between socio-economic status of villages and risk of brucellosis. Most of the endemicity Villages supplied with clean water and electricity had a was seen in the north and south villages of Bardsir which twofold increase in the probability, compared to villages had more density of cattle (No. per square kilometres). without supplies. Moreover, villages with health facilities had four times greater risk (31.8% compared to 7.5% in village without health facilities). Villages with higher literacy (50% and above) and villages located at flatlands had higher risks (16.4% compared to 7.8% in villages with lower literacy), but the statistical tests revealed that the differences were not significant (Table 2). Using GIS, it seems that we may generate better conclusion on human health in Iran if we link the human and veteran data. It is clear that we can implement more effective control programme if we use the data of both human and animals. Both systems are collecting data and generating comprehensive datasets; therefore, we just need to create a link to combine whole information. overall it seems that, it does not add much more cost to the ongoing systems, but it could generate much more efficient results. According to our findings in this study, it is not difficult to explain the associations between the human and livestock populations and risk of brucellosis. Since cows

and sheep are the main sources of the disease these facilities or because of great human and livestock transmission to human, the number of effective contacts population in these villages. would be a result of livestock and human population sizes which was also observed in other studies.[3, 6]. However, we should acknowledge to the limitations of such a model. We used surveillance data which are not In Zinstagg et.al survey on livestock-to-human perfect from the research point of view. Nevertheless, in brucellosis transmission in Mongolia in a period of 10 Iran such as other endemic areas, the numbers of reported years (1990-99), it was shown that trends of brucellosis in brucellosis are always underestimated. The low incidence both human and animals were identical [7]. They reported in known brucellosis-endemic areas may reflect mentioned more domestic infected animals, increased the the absence or the low levels of surveillance and risk of human brucellosis, which is similar to our finding. reporting[9]. We have to mention that we produced a Although they mentioned that more than 90% of human model based on ongoing data to check its effectiveness in brucellosis in Mongolia was small-ruminant (sheep) real system. Furthermore, we did not have accurate derived, but it seems that brucellosis in Bardsir is mostly village based information on the animal vaccination. cattle derived. Which is not compatible with the fact that These types of information can be collected in veteran most common species is brucella melitensis in Iran[8]. system if their staffs know the importance of such type of data. Furthermore, despite performing control programme and For better conclusion, further investigation on other livestock's vaccination more than one decade, annual possible risk factors, e.g. disease prevalence of animal incidence of human brucellosis in rural areas of Bardsir population, livestock vaccination coverage and local remain considerable. It seems that such a control customs would give some new evidence of local animal programme in rural areas could not be as effective as population association with epidemiology of human urban. Perhaps, this is partly due to difficulties in disease. Also we suggest that more study should be conducting thorough and sustainable disease control carried out to explore the feasibility of GIS systems on programme, and partly due to the wide and scattered the other types of common diseases between human and distribution of livestock in rural areas which cause more animals in larger scales. effective contact with infected animals and their diary products. In this study, we did not focus on determining As conclusion, this study not only showed the implication each risk factor of brucellosis, but some local customs, of GIS to link human and animal data, but also exposed environmental variation, difference in species of the limitations and weak points in ongoing systems. pathogens and their distribution are likely to be factors Therefore, we suggest that the impact of control that would affect the local incidence differences .it's programmes can be improved by using GIS methods and believed that fresh cheese and milk are the major source linking human and animal data. of human brucellosis in Iran, and the disease is more common in rural areas, where cheese is usually produced by local small factories or in households[6]. However, it is difficult to estimate to what extent each of the two transmission routes (occupational or food-born) takes part in the human disease incidence in rural area, as both could be deeply related to the size of local animal population. In contrast with our expectation, access to satiated water, electricity and health facility and even asphalt road increased the risk of brucellosis in our univariate models. This could be either because of a more accurate diagnose and reporting system in those villages with access to

Table 1 Frequencies of villages with at least one positive case, classified by potential risk factors, in Bardsir district. Villages with at least on case (%)   Variable   Number of Number (%) 95% Confidence villages 34 (10.9) interval (%) Number of inhabited villages 311 7.4 - 14.4 Human population 1-10 100 5 (5) 0.7 - 9.3 (person) 11-50 118 6 (5.1) 1.1 - 9.1 51-100 39 4 (10.3) 0.3 - 20.2 >100 54 19 (35.2) 22.0 - 48.3 Animal population 0-9.9 67 4 (6.0) 0.1 - 11.8 Sheep/goat 10-99.9 58 3 (5.2) 0.0 - 11.0 (heads per 100-499.9 115 19 (16.5) 9.6 - 23.4 Sq. km) >=500 71 8 (11.3) 3.7 - 18.8 Cattle 0-0.9 141 6 (4.3) 0.9 - 7.6 (heads per 1-9.9 56 5 (8.9) 1.2 - 16.6 10-19.9 52 5 (9.6) 11.2 - 26.8 Sq. km) >=20 62 18 (29.0) 1.5 - 55.6 Socio-economic factors no 183 12 (6.6) 3.0 - 10.1 Access to yes 128 22 (17.2) 10.7 - 23.7 piped water Access to no 131 8 (6.1) 2.0 - 10.2 electricity yes 180 26 (14.4) 9.3 - 19.6

Access to no 267 20 (7.5) 4.3 - 10.6 health facility yes 44 14 (31.8) 18.1 - 45.6 Asphalted road no 262 19 (7.3) 4.1 - 10.4 yes 49 15 (30.6) 17.7 - 43.5 Education <49 64 5 (7.8) 1.2 - 14.4 Literacy (%) >=50 146 24 (16.4) 10.4 - 22.5 Geographic factors Location of mountain 165 13 (7.9) 3.8 - 12.0 146 21 (14.4) 8.7 - 20.1 village flatlands * chi-square test for trend  Table 2. Crude and adjusted odds ratios of potential risk factors for endemic villages, only

human and animal populations were entered in multivariate analysis Crude adjusted Variable   OR 95% CI OR 95% CI Population* ** 2.27 1.68-3.06 1.94 1.41-2.67 Log(population) 1.74 1.35-2.26 1.81 1.18-2.78 Log(cattle density+1) 1.16 0.97-1.39 0.75 0.56-1.01 Log(sheep density+1) Socio-economic indicators 2.96 1.39-6.30 - - Access to water 2.60 1.13-5.99 - - Access to electricity 5.76 2.56-12.99 - - Access to health facility 5.64 2.54-12.53 - - Asphalted road Education 2.32 0.84-6.45 - - High literacy (>50%) Geographic factor   Located in flatlands 1.96 0.94-4.10 - - * Observed variable converted into its natural logarithm. ** Univariate logistic regression was used to estimate odds ratios for these continuous variables (odds ratios for 1 unit increase in the explanatory variable) REFERENCES 2.Asadi-Lari M, et al., Public health improvement in Iran--lessons from the last 20 years. Public Health, 2004. 1.Shadpour K., Primary health care networks in the 118: p. 395-402. Islamic Republic of Iran. East Mediterr Health J, 2000. 6: p. 822-5. 3.Pappas G., et al., medical progress; Brucellosis. N Engl J Med, 2005. 352: p. 2325-36.

4.Moore DA and Carpenter TE, Spatial analysis methods and geographic information systems: use in health research and epidemiology. Epidemiol Rev, 1999. 21: p. 143-61. 5.Statistical Centre of Iran, Statistical Yearbook 1382 (March 2003 - March 2004). 2005, tehran: Statistical Centre of Iran Department of Publication and Information. 6.Almuneef MA, Memish ZA, and Balkhy HH, Importance of screening household members of acute brucellosis cases in endemic areas. Epidemiol Infect, 2004. 132: p. 533-40. 7.Zinsstag, j., et al., A model of animal–human brucellosis transmission in Mongolia. Preventive Veterinary Medicine, 2005. 69: p. 77-95. 8.Panahi M, Brucellosis, in Epidemiology and control of common disorders in Iran. 2003, Khosravi publication company: Tehran. p. 533-41. 9.Jacques godfroid, et al., From the discovery of the Malta fever’s agent to the discovery of a marine mammal reservoir, brucellosis has continuously been a re- emerging zoonosis. Vet. Res., 2005. 36: p. 313-326.

Technical Session - 5 HealthGIS Database Critical Factors Influencing Electronic Health Records (EHR) Implementation: Important Lessons for Saudi Arabia Public Hospitals Fares Alshammari and Shane Thomas…………………………………………………………………….111 Survey about PKU Patients Follow up and Prevalence from 2006 to 2008 at Tehran University's Region Azita Karimi, Fatemeh Haji Ali Asgari, Hamid Khodabandehloo Soltani and Mohammad Ali Shokri……………………………. ………………………………………………............112 GIS - Health Infrastructure Mapping for Andhra Pradesh State Rajiv Sharma I.A.S , K. Dakshina Murthy and N. Satyanarayana……………………………………..…115 Open Source GIS Based Framework for Development of Health SDI R. D. Gupta, Arun B. Samaddar, R. K. Barik and Marguerite Madden …………………………………119 Heath GIS for Part of Trivallore Town using Cartosat1 & LISS 4 Pan Merged Data S. Mohamed Ghouse, J. Rathika, V. Balamurugan and V. Ramakrishnan………………………………124 Building a Georeferenced Population Health Registry Seetharaman Narayanan, Chandrasekaran Ramaswamy , Christopher Amalraj Vallabadas, Hari Srimathi, Shanmuganathan Pasupathy and Vadivu Ganesan.………..…………………………….127 Investigation of Tuberculosis Clusters in Dehradun City using Geographical Information System and Spatial Scan Statistic Neeraj Tiwari, K.Ram Mohan Rao and V. S. Tolia……….……………………………………………….130

Critical Factors Influencing Electronic Health Records (EHR) Implementation: Important Lessons for Saudi Arabia Public Hospitals Abstract Fares Alshammari and Shane Thomas It has become clear that paper record systems are unable to meet contemporary patient information needs and that well implemented and designed electronic health records (EHRs) can do so. Yet, despite this situation it has been shown internationally that the implementation of EHRs is low and that there are significant barriers to EHR implementation. Aim: The aim of this paper is to explore the influence of hospital managers on EHRs implementation in their hospitals. This paper is based on a comprehensive review of the research literature to define the relevant factors that influence EHR implementation in order to produce useful guidelines for EHRs implementation. Results: The results show that a range of barriers to and facilitators of EHR implementation have been identified. These include issues relating to IT technical requirements and infrastructure but also change management, organizational policies and the motivational and training preparation of the management and workforce to embrace the changes required by EHR implementation as well as their involvement in the actual change process in healthcare delivery and healthcare workers expectations and involvement. This paper using an evidence-based review identifies the key factors that influence successful EHR implementation that should be identified before costly investments are made. The paper discusses the implications of these findings for the implementation of EHR systems into MOH public hospitals in Saudi Arabia.

SURVEY ABOUT PKU PATIENTS FOLLOW UP AND PREVALENCE FROM 1386 TO 1387 AT TEHRAN UNIVERSITY S REGION. Azita Karimi, Fatemeh Haji Ali Asgari, Hamid Khodabandehloo Soltani and Mohammad Ali Shokri Tehran University of medical sciences, Tehran, Iran [email protected],[email protected], [email protected] ABSTRACT: Introduction: It is a congenital disease (autosomal recessive) and defined as plasma phe. Above 2mg/dl. The most important manifestation is mental retardation .although infants appear NL at birth. Material & type: it was descriptive cross sectional study, which data gathered from PKU society of Iran and new born screening program from centers of this university .Finding: there are 126 patients in Tehran university region from about 1200 patients recognized in all of the country. 62 girls (49.21%) & 64 boys (50.80%) Average age of diagnosis: 18.9 months. Most common presentation: developmental delay. 3 patients got high school diploma and one entered college. 20%attended regular schools but quitted in the middle school, 15.7% attended school for retarded children, and 64% never attended school. 81% of the parents were first cousins.11 families had more than one pku child and 2 families had 3 children affected. Results: new born screening is the best method for early detection of hyperphenylalaninemia. Cases treated in the first month have higher IQ (mean 95) than treated between 31-65 days (mean 85). Famial marriage is very important too. KEY WORDS: IQ, screenhng, familial marriage, epidemiology 1. INTRODUCTION or mousy\" odor of skin, hair, sweat and urine (due to phenyl acetate accumulation); and a Phenylketonuria (PKU) is an autosomal tendency to hypo pigmentation and eczema are recessive genetic disorder characterized by a also observed. In contrast, affected children deficiency in the enzyme hepatic who are detected and treated are less likely to phenylalanine hydroxylase. develop neurological problems and have seizures and mental retardation, though such Screening and presentation :Blood is taken clinical disorders are still possible. from a 3-5 days old infant to test for Path physiology phenylketonuria PKU is normally detected Classical PKU is caused by a mutated gene for the enzyme phenylalanine hydroxyls (PAH), using the HPLC test, but some clinics still use which converts the amino acid phenylalanine to other essential compounds in the body. the Guthrie test, part of national biochemical Other, non-PAH mutations can also cause PKU. screening programs. Most babies in developed Classical PKU countries are screened for PKU soon after The PAH gene is located on chromosome 12 in birth. If a child is not screened during the the bands 12q22-q24.1. More than four hundred disease-causing mutations have been routine Newborn Screening test, the disease found in the PAH gene. PAH deficiency causes a spectrum of disorders including may present clinically with seizures, albinism classic phenylketonuria (PKU) and hyperphenylalaninemia (a less severe (excessively fair hair and skin), and a \"musty accumulation of phenylalanine). KU is an autosomal recessive genetic disorder, meaning odor\" to the baby's sweat and urine (due to that each parent must have at least one mutated phenylacetate, one of the ketones produced). In most cases a repeat test should be done at approximately 2 weeks of age to verify the initial test and uncover any phenylketonuria that was initially missed. Untreated children are normal at birth, but fail to attain early developmental milestones, develop microcephaly, and demonstrate progressive impairment of cerebral function. Hyperactivity, EEG abnormalities and seizures, and severe learning disabilities are major clinical problems later in life. A \"musty

allele of the gene for PAH, and the child must different human populations from 1 in 4,500 inherit two mutated alleles, one from each births among the population of Ireland to parent. As a result, it is possible for a parent fewer than one in 100,000 births among the with PKU phenotype to have a child without population of Finland. PKU if the other parent possesses at least one functional allele of the PAH gene; but a child 2. HISTORY ABOUT MEDICAL of two parents with PKU will always inherit SCIENCE OF TEHRAN UNIVERSITY two mutated alleles, and therefore the disease. The first modern center for medical training in Tetrahydrobiopterin-deficient Iran was founded in 1851. It was a part of the hyperphenylalaninemia Institute for Higher Education (Dar-ol- Fonoon). The School of Medicine was A rarer form of the disease occurs when PAH established as a part of the University of is normal but there is a defect in the Tehran in 1934. It is the oldest and the most biosynthesis or recycling of the cofactor outstanding medical center in the I.R. of tetrahydrobiopterin (BH4) by the patient. This IRAN; · is one of the country's top research cofactor is necessary for proper activity of the universities, receiving an annual grant of over enzyme. 300 billion Rials from the government; · accepts applications from only the most Metabolic pathways. qualified Iranian high school graduates who have already passed the National Entrance Phenylalanine is a large, neutral amino acid Exam and very top students from abroad; · has (LNAA). LNAAs compete for transport across more students in medical based courses than the blood-brain barrier (BBB) via the large any other higher education institution in Iran. neutral amino acid transporter (LNAAT). It is covered south of Tehran, eslamshahr and Excessive phenylalanine in the blood saturates ray . the transporter. Thus, excessive levels of phenylalanine significantly decrease the levels Supplementary infant formulas are used in of other LNAAs in the brain. But since these these patients to provide the amino acids and amino acids are required for protein and other necessary nutrients that would otherwise neurotransmitter synthesis, phenylalanine be lacking in a low phenylalanine diet. These accumulation disrupts brain development, can continue in other forms as the child grows leading to mental retardation. up such as pills, formulas, and specially formulated foods. (Since phenylalanine is Treatment necessary for the synthesis of many proteins, it is required but levels must be strictly If PKU is diagnosed early enough, an affected controlled). In addition, tyrosine, which is newborn can grow up with normal brain normally derived from phenylalanine, must be development, but only by eating a special diet supplemented.) low in phenylalanine for the rest of his or her life. This requires severely restricting or 3. MATERIAL EQUIPMENT eliminating foods high in phenylalanine, such as meat, chicken, fish, nuts, cheese, legumes The screening was present in all centers of and other dairy products. Starchy foods such as neonatal screening from 1384 (2005). Tums potatoes, bread, pasta, and corn must be University has 11 screening centers for monitored. Infants may still be breastfed to neonates. Our data gathered from 3-5 days provide all of the benefits of breast milk, neonates (from hills) of these centers. Now in though the quantity must be monitored and Iran are diagnosed gene defects with genetics supplementation will be required. Many diet exam for patients & their parents. There is do foods and diet soft drinks that contain the abortion if it s defect is diagnosed in other sweetener aspartame must also be avoided, as pregnancies and there is genetic consultation aspartame consists of two amino acids: for familial marriage too. phenylalanine and aspartic acid. 4. CONCLUSION Incidence The maps are prepared using GIS to visualize The incidence of PKU is about 1 in 15,000 the situation of one important genetic disease, births, but the incidence varies widely in PKU. The world is fighting to eradicate it.

Many countries have screening programs for public health system in the developing detecting that. The use of GIS has shown the country. It is suggested that the similar study is impact zone .GIS helps in analyzing the done for the larger area would become more available information .as a result the system useful and effective. can become more efficient, productive and capable to make up the future tasks very efficiently .this study is a representative of any REFERENCES tetrahydrobiopterin): a phase II, multicentre, open-label, screening study\". 1. ^ Center wall, S. A. & Center wall, W. R. Journal of Inherited Metabolic Disorders (2000). \"The discovery of 30: 700–707. doi:10.1007/s10545-007- phenylketonuria: the story of a young 0605-z. PMID 17846916. couple, two affected children, and a http://www.nature.com/nrd/journal/v7/n3/f scientist.“ Pediatrics 105 (1 Pt 1): 89– ull/nrd2540.html. 103. doi: 10.1542/peds.105.1.89. PMID 6. ^ Guldberg, P., Henriksen, K. F., Sipila, 10617710. I., Guttler, F., de la Chapelle, A. (1995). http://pediatrics.aappublications.org/cgi/co \"Phenylketonuria in a low incidence ntent/full/105/1/89. population: molecular characterization of mutations in Finland\". J. Med. Genet 32: 2. ^ Piet, J., Kris, R., Rupp, A., Mayatepek, 976–978. doi:10.1136/jmg.32.12.976. E., Rating, D., Boesch, C., Bremer, H. J. PMID 8825928. (1999). \"Large neutral amino acids block phenylalanine transport into brain tissue in patients with phenylketonuria\". Journal of Clinical Investigation 103: 1169–1178. doi:10.1172/JCI5017. PMID 10207169. http://www.jci.org/cgi/content/full/103/8/1 169. 3. ^ Burton, BK; Kar S, Kirkpatrick P (2008). \"Fresh from the Pipeline: Sapropterin\". Nature Reviews Drug Discovery 7: 199–200. doi:10.1038/nrd2540. http://www.nature.com/nrd/journal/v7/n3/f ull/nrd2540.html. 4. ^ Michals-Matalon K (2008). \"Sapropterin dihydrochloride, 6-R-L- erythro-5,6,7,8-tetrahydrobiopterin, in the treatment of phenylketonuria\". Expert Opin Investig Drugs 17 (2): 245–51. doi:10.1517/13543784.17.2.245. PMID 18230057. 5. ^ Burton BK, Grange DK, Milanowski A, Vockley G, Feillet F, Crombez EA et al. (2007). \"The response of patients with phenylketonuria and elevated serum phenylalanine to treatment with oral sapropterin dihydrochloride (6R-

GIS - HEALTH INFRASTRUCTURE MAPPING FOR ANDHRA PRADESH STATE Rajiv Sharma I.A.S1, K.Dakshina Murthy2, N.Satyanarayana3 1Director General, 2Project Manager, 3Team Leader (GIS), Centre for Good Governance, Hyderabad, India ABSTRACT: GIS is becoming a vital tool in healthcare applications covering database management, planning, risk assessment, service area mapping, location identification etc. One of the reasons for sudden surge of GIS usage in healthcare planning is the spatial dependency of health related factors and limited resources for ever increasing demand. The GIS technology helps the policy makers in better planning and also optimum utilization of available health resources. The GIS program aims to add value and enhance the capabilities of the HM&FW department for providing consistent, timely, accurate information to support the common decision making activities. With the help of this web application system it is easy to monitor the heath infrastructure across the state, up to the village level. The spread of disease in a particular area and availability of infrastructure facilities in the same area are brought together at the same level to find out critical areas demanding immediate attention. KEY WORDS: Health GIS, HRIMS, Health, Infrastructure, HM&FW. 1. INTRODUCTION where reform initiatives are implemented is the department of Health. “Public health management needs information on various aspects like the prevalence of diseases, facilities This programme encouraged reforms in Health that are available in order to take decisions on either department through Strategy and Performance creating infrastructure facilities or for taking immediate Innovation Unit (SPIU) created within the department. action to handle the situation and so on. These decisions One of the major projects taken up by the SPIU was to need to be taken based on the observations made and have a GIS enabled platform. It aims to add value and available data. As the data relates to Public health enhance the capabilities of the HM&FW department for covering the whole state and the entire population the providing consistent, timely, accurate information to data is voluminous, and hence it is extremely difficult to support the common decision taking activities of the understand the real content. The data needs to be office as it strives to meet the state government goal of presented in a way that the temporal and spatial nature equitable healthcare accessible tor all the people of of the problem can be brought out in a focused way”.[1] state. “The availability of statistical and other information in By using GIS, the department will be able to map health spatially referenced form and the functions provided by facility based on serviceable area, proximity to available a GIS could allow analyses that were previously too facilities and population distribution, identify the expensive or impossible to perform. Geographic distribution of resources and discrepancy of service Information System (GIS) is an innovative technology, distribution in spatial domain, recognize the deprived ideal for generating data suitable for analysis both with areas and locate areas for future investments based on respect to space and time.” [2] accessibility and social status. The Department of Health, Medical and Family Welfare 2. OBJECTIVES provides health care facilities to the people of Andhra Pradesh. The role of the department is defined as The objectives of this GIS supported initiatives are as follows: “HM & FW constructs primary health centres, follows: sub-centres, hospitals, dispensaries, clinics and other  Creating Health infrastructure GIS spatial database health care centres. It undertakes schemes for acquisition of medical equipment and implementation of of all the departments and sub departments under infrastructure facilities; improve the functioning of Health Medical and Family Welfare for Health GIS hospitals etc.”[3] web Application.  To monitor and analyze the health facilities and The Government of Andhra Pradesh as part of its their serviceable area, proximity and their distribu- governance reform initiatives took up a project named tion across the state, up to the village level. “Andhra Pradesh Public Management and Service  To analyze the disease pattern in a particular area Delivery Improvement Programme” with the assistance and availability of infrastructure facilities in the of DFID during 2006-09. One of the key departments same area together, to find out critical areas de- manding immediate attention.

Mother and Child Hospitals 3. HEALTH GIS SPATIAL DATA Paediatrics The Health Medical and Family Welfare department has 6 Andhra various departments and sub departments under it. The location based spatial information for each Pradesh sub departments are mapped at village level for each department. The attribute information for each facility Health & is also added as given by the sub department nodal officer and got verified respectively for each department. Medical - The system follows the hierarchy of the entire Health Department and maps it to their respective Housing & databases in a synchronized manner. The details of official hierarchy to exercise managerial control as Infrastructur well as the man power resources are also recorded into the system. The department wise health e facilities marked under each sub departments are as follows [4]: Developmen t Corporation (APHMHID C) 7 Drug Control - Administrati on (DCA) 8 Commission Comprehensive Emergency er of Family Obstetric & Neonatal Care Centre Welfare (CEMONC) (CFW) Post Partum Unit Sub Centres S.No Departme Sub Departments Urban Family Welfare Centres 1 - nt Urban Health centre Institute of 9 Ayurvedic, Ayurveda Preventive Medicine Yoga & (IPM) Naturopathy, Unani Unani, Siddha & Homeopathy 2 Andhra ICTC – Integrated Counselling and Homeopathy Pradesh State Testing Centres (AYUSH) Aids Control STD Clinics-Sexually Transmitted 10 Andhra - Society Disease Clinics Pradesh (APSACS) Yogadhyaya BB- Blood Banks na Parishad ACSR-AIDS Case Surveillance (APYP) Centres ART Centres-Anit Retroviral 4. HEALTH GIS WEB APPLICATION Therapy Centres The Spatial data of Health department is used further in C&S centres – Care & Support developing a web application using ESRI products such as Centres ARCIMS and ARCSDE [5]. The web application gives information on various health facilities of each NGO-TI–NGO Targeted sub departments under every department. The application is Interventions secured by proper authentication and login access. The login facility is given to each Heads of Department 3 Directorate Directorate of Medical Education and Secretaries etc. of Medical Education DME Hospitals (DME) DME Nursing Schools DME Nursing Colleges 4 Directorate Civil Hospitals of Health (DH) Community Health Centre Dispensaries Mobile Medical Unit 5 Andhra Primary Health Centres Area Hospitals Pradesh Vaidya Civil Hospitals Vidhana Parishad Chest Disease Hospitals (APVVP) Community Health Centre District Hospitals

Health Facilities Information: Database Design [3] The user is provided with tools for extracting information of each facility, such as identity, selection apart from the basic GIS tools like Zoom and Pan. Buffer analysis tool for proximity analysis and drill- down functionality is also provided upto to village level. Help is provided for the usage of tool. The user manual describing the various flow of application and its usage is also provided for the users. Online editing facility to update the attribute data is also provided in order to update the information. Health Facilities Information: PHC’s across A.P. 3. GIS TOOLS The following figures show the general look of the web application and its usage. Health Facilities Information : GIS Query Builder Health Facilities Information Health Facilities Information : Query Results Health Facilities Information: Blood Banks across A.P.

PHC : Primary Health Centre SPIU : Strategy and Performance Innovation Unit ACKNOWLEDGEMENT Buffer Showing PHC within 10km radius We would like express our gratitude to all those who from DME Hosiptals made it possible to complete this application. We would like to thank Sri L.V.Subrahmanyam, Principal Secretary (Health) Andhra Pradesh for his support as well as the then Joint Secretary & Director (SPIU), HM&FW Dept. Smt. Shailaja Ramaiyer, for initiating the project. We would like to thank Dr. Mohd. Ariz Ahammed, Addl. Secretary, HM&FW Dept. in refining and streamlining the project. Our thanks are also due to CGG staff, who have helped in various ways by spending time in developing the application. 4. CONCLUSION REFERENCES The Health Medical and Family Welfare department of 1. http://www.gisdevelopment.net/application/ Andhra Pradesh has huge infrastructure facilities under health/overview/healtho0006pf.htm every department. The distribution and functionalities of each department and their sub department should be 2. http://www.gisdevelopment.net/application/lis/ brought together to analyze the health facilities in a rural/lisr0008.htm serviceable area with proximity to their respective villages. The web based application aims to enhance the 3. http://www.aponline.gov.in/apportal/depart- capabilities of the HM&FW department for providing ments/departments.asp?dep=16&org=98 consistent, timely, accurate information to support the common decision making activities. 4. Andhra Pradesh Health Medical & Family Wel- fare’s reference materials on each sub depart- This also helps in planning infrastructure, to create new ments of every department and their respective facilities based on gap analysis which can be determined websites. on the basis of the spatial distribution of existing facilities. The application may further be enhanced and 5. ESRI product documentation & training manu- integrated with MIS, which may further help in als on ARCIMS and ARCSDE analyzing different parameters such as disease spread, disaster management under major health calamities, providing the proximity of facility from an accident area or emergency services. ABBREVIATIONS ARCIMS : Arc Internet Map Server ARCSDE : Arc Spatial Database Engine CGG : Centre for Good Governance DFID : Department for International Development ESRI : Environmental Systems Research Institute GIS : Geographical Information System HM&FW : Health, Medical and Family Welfare HOD : Head of the department HRIMS : Health Resources Information Management System MIS : Management Information System

OPEN SOURCE GIS BASED FRAMEWORK FOR DEVELOPMENT OF HEALTH SDI R.D. Gupta 1, Arun B. Samaddar 2, R.K. Barik 3, Marguerite Madden 4 1- Professor & Coordinator, GIS Cell; Department of Civil Engineering; Motilal Nehru National Institute of Technology, Allahabad- 211004, India (e-mail: [email protected]) 2- Director, Professor & Chairman, GIS Cell, Motilal Nehru National Institute of Technology, Allahabad- 211004, India (e-mail: [email protected]) 3- M. Tech. (GIS & Remote Sensing); GIS Cell; MNNIT, Allahabad, India (e-mail: [email protected]) 4- Professor, Centre for Remote Sensing and Mapping Science, Department of Geography, University of Georgia, Athens, Georgia- 30602-2503, USA (e-mail: [email protected]) ABSTRACT: Spatial Data Infrastructure (SDI) is a portal where each stakeholder can access, use and exchange spatial data for social, economic and environmental activities and is increasingly being acknowledged as a national resource. GIS is a valuable tool to assist in health education and research as well as to plan, monitor and evaluate various health programmes and health systems. Disease phenomena are associated with spatial and temporal factors and web enabled GIS can provide a real time and dynamic way to represent disease information through maps and statistics. Several countries have taken initiative to establish a well organised GIS based public health information infrastructure. In India, there is a need for development of Health SDI at national level with its nodes being at State level, district level and then at village level. In the present work, an open source GIS based web enabled framework for Health SDI implementation has been developed. The proposed framework is distributed and interoperable, and uses thin-client architecture. The open source GIS software used for development of Health SDI framework include Open JUMP GIS for creation of geospatial health database, MYSQL for storing of security aspects and non-spatial data. ALOV and Apache Tomcat for imparting geospatial web capabilities in terms of Web Map Services (WMS), Web Feature Services (WFS), Web Coverage Service (WCS); and PHP: Hypertext Preprocessor for dynamic server side scripting. KEY WORDS: Open Source GIS, Health SDI, Web GIS 1. INTRODUCTION The available open source GIS software can be used for the works related to database creation, spatial modelling Now-a-days, open sources GIS software can provide the and web based services (Barik et al., 2009). users an alternative and competitive way of modelling the real world problems to produce robust spatial Spatial data are crucial for the development of a region solutions in comparison to costly proprietary GIS and there is increased recognition for the benefits being software. In a proprietary market, the closed source returned to communities by investing in the codes are providing for the needs of industries and development and implementation of spatial information organisations such that the users can access various systems. The nations have started making unprecedented methods built in the software, however, the working of investments in data and the means to produce and use it. these methods remain hidden. The proprietary software These countries believe they can be benefited both is sometimes too costly for small organisations to run economically and environmentally from better their business in cost effective manner. Camara (2004) management of their spatial information by taking a emphasized that Open Source Software (OSS) projects perspective that starts at a local level and proceeds can offer rich functionality, robustness, cooperation through state and national levels to a global level. This from contributing developers and continuous has resulted in the development of the Spatial Data improvement. Infrastructure (SDI) concept with hierarchical relation at these different levels. The sharing of data through web is Fortunately, there are many open source software that one of the prime issues in SDI. can compete the proprietary software in the field of GIS. Programmers have created several open source libraries SDI is an initiative intended to create an environment in and GIS suites to cope with the flood of GIS data and which all stakeholders can cooperate with each other their formats. The goal of Open Geospatial Consortium and interact with technology to better achieve their (OGC) is to encourage the use of open source GIS objectives at different political/ administrative levels. standards and development of community-led projects. SDI has become very important in determining the way

Health GIS 2009 in which spatial data are used throughout an SDI concept that is depicted in Figure 1 (Mansourian et organisation, a nation, different regions and the world. al., 2005). In principle, SDI allows the sharing of data, which enables users to save resources, time and effort when Dynamic trying to acquire new datasets by avoiding duplication of expenses associated with generation of new data (GSDI, PePoepolpele Accessing HDHeDaaetlaaatthlath 2004). Network For health sector, disease data sharing is important for Policy the collaborative preparation, response, and recovery stages of disease control. Disease phenomena are Standards strongly associated with spatial and temporal factors. Web-based GIS provides a real-time and dynamic way Figure 1: The Dynamic Nature of SDI Model to represent disease information on maps. However, data heterogeneities, integration, interoperability, and The development of SDI at national level in India, i.e., cartographical representation are still major challenges Indian NSDI was initiated in 2000 jointly by Department in the health geographic fields (Vanmeulebrouk et al., of Science and Technology (DST) and Indian Space Re- 2008). These challenges cause barriers in extensively search Organisation (ISRO) through the establishment of sharing health data and restrain the effectiveness in a national task force to prepare an action plan under the understanding and responding to disease outbreaks. To aegis of DST (Puri et al., 2007). A number of govern- overcome these challenges in health sector, mapping and ment organisations like Survey of India, Geological Sur- sharing of spatiotemporal disease information in an vey of India, Department of Space, Forest Survey of In- interoperable framework based on OGC specifications dia, etc. are involved in its creation. under GIS environment is the need of the hour. In developed countries, SDI has become a part of the 2. ROLE OF SDI basic infrastructural facilities that is efficiently coordinated and managed for development. Several The spatial data communities participate in different countries have also taken initiative to establish a well levels of decision-making, from strategic to tactical to organised GIS based disease mapping and surveillance operational, in the management and utilisation of spatial system and public health information infrastructure. In data. The development of SDI supports these decision- India, there is a need for development of Health SDI at making functions at different administrative and national level with its nodes being at State level and political levels. The influence of SDI through decision- then at District level. The district level node of Health making frameworks varies in accordance with the SDI will cover the information at micro level, i.e., jurisdiction, the differing levels of decision-making village level. The development of Health SDI can be functions as well as the support of the spatial taken up as the policy matter by the Indian Government community. SDI provides an environment within which and can be implemented as one of sub-systems under organisations interact with technologies to foster Indian NSDI. activities for using, managing and producing geographic data. It is the technology, policies, standards, human 3. OBJECTIVES OF THE PRESENT WORK resources, and related activities necessary to acquire, process, distribute, use, maintain and preserve spatial The present study deals with the development of web data (Rawat, 2003). based framework for implementation of Health SDI by using open source GIS software. The framework mainly The core components of SDI can be viewed as policy, focuses on publishing maps, geospatial health data, networking, standards, people and data. These can be sharing of metadata and geospatial web services to grouped into different categories based on the nature of develop prototype SDI for health sector. their interactions within the SDI framework (Rajabifard et al., 2002). The people and data can be taken as one Allahabad district of Uttar Pradesh State in India has category while the second category can be the main been taken up as the study area to demonstrate the technological components, i.e., the access network, functionalities of developed Health SDI framework. In policy and standards. The nature of the second category particular, Malaria, a vector borne disease has been is very dynamic due to the rapidity with which taken up for investigations. The district has an area of technology develops. This suggests that an integrated 7261 sq. km and lies between latitudes 24o47'00N and SDI can not be thought as having spatial data, value- added services and end-users alone, but instead involves other important issues regarding interoperability, policies and networks also. This reflects the dynamics of

Health GIS 2009 25o47'00N and longitudes 81o09'00E and 82o21'00E for storing of security aspects and non-spatial data for with total population of about 50 lacs. decision making. PHP: Hypertext Preprocessor language has been used for dynamic server side scripting in the 4. FRAMEWORK OF HEALTH SDI framework. For development of Health SDI framework, the main One important vector-borne disease, i.e., Malaria has focus has been on the use of a practical approach to been selected to demonstrating the capabilities of explore and extend the concept of SDI for health sector. developed framework for Allahabad district. In The framework provides an effective and efficient means particular, the information in terms of the various for sharing geospatial health data and non-spatial data aspects of Malaria, i.e., susceptibility regions, Annual and for publication of maps on the web using GIS in a Parasite Index (API), Annual Blood Smear Examination secure way. Figure 2 shows the flow of information for Rate (ABER) and Slide Positive Rate (SPR) have been the Health SDI framework developed in the present taken for prototype development. The user can add more work. information about Malaria or any other disease at a later stage. Geospatial Health Database 5. HEALTH SDI: PROTOTYPE DEVELOPMENT (Using Open JUMP) The prototype development is based on Jacobson’s method Storing of Geospatial Non-spatial data of Object Oriented Software Engineering (OOSE) for health data in shape file incorporating strong user focus and the time critical nature (Mall, 2004). The OOSE method involves the formation of Business Process Data models that capture the actors of the system and their behaviour for each of the design stages. The models are Web Feature made up of objects representing real world entities. This is Services a natural way for people to describe their environment; and helps in reducing the semantic gap between the developed model and the real world. Figure 3 shows the complete process model for development of Health SDI framework. Web Map ALOV, Apache Tomcat, Utility Services MySQL & PHP Services Web Coverage Services Client 1 Client 2 ......... Client N Figure 3: Process Model for Health SDI Figure 2: Flow Chart showing Health SDI Framework The software development process adopts a sequence of steps including requirement specification, analysis and For the creation of integrated geospatial health database, design, implementation and testing, complete module and the present framework uses Open JUMP GIS software. model observation. The process is usually cyclic or ALOV and Apache Tomcat have been integrated for incremental in nature and each implementation refines the imparting the geospatial web capabilities with respect to analysis and design stages through evaluation and testing of Web Map Services (WMS), Web Feature Services a completed module. This successive version development (WFS) and Web Coverage Services (WCS). Geospatial allows to take into consideration more informed view of web capabilities indicate to a web based GIS which can SDI requirements. Further, the incremental development be modelled using the client-server architecture. The strategy allowed the problem of constructing this business process data are meant to give additional framework to be tackled in smaller, more manageable storage for security purpose and for business application. portions of increasing complexity. In addition, it is A thin client model is used where most of the processing expected that each module would reveal unique features work is done on demand in the server. MySQL is used related to the requirements of the underlying

Health GIS 2009 infrastructure and enable exploration of the interfaces Further, in the design phase, the flow of processes in the between SDI components. system is captured in the form of state diagram which need to be properly designed. In context of the present 5.1 Requirement Specifications work, state diagram describes the workflow behaviour of Health SDI and is shown in Figure 5. The requirement stage of application design aims to specify the behaviour of the framework from the Figure 4: State Diagram for Health SDI perspective of a user. Understanding potential system 5.3 Implementation and Testing users in terms of knowing who they are and what they In this phase, coding is done to fulfil the overall want to do facilitate system development. The use case requirements of the framework and for testing the model is one of the important aspects in this phase appropriate parameters. The planning and design which basically specifies a sequence of actions that need parameters are kept in mind during coding. In the to be performed. The construction of use cases involves prototype development, coding is one of the challenging how a user would interact with the system to identify the tasks for achieving the real success. The coding is various system objects and highlights the need for performed in the PHP environment and follows the careful consideration to be given to the issues of human incremental option. The testing phase is very essential computer interaction within the network environment. for increasing the QOS (Quality of Service) of the software. The testing has been carried out in accordance In the present work, the use case model has been to the iterative model after developing code for whole associated with three types of users, i.e., administrative process. user, general user and developer. The administrative 5.4 Complete Module user will have the authority to view, delete and find The framework consists of three main modules, e.g., existing users which are associated with this system. The module I for registration, module II for Allahabad general user has the variety of option like login, logout, District Malaria Mapping, and module III for Utility register, get-information, maximize map, minimize map, Services. Module I describes the detailed process to pan feature, get-coordinate, upload files and download files. The developer is associated with the system in terms of user tools, modifying the existing system and collecting the feedback on the system. 5.2 Analysis and Design Translating the user requirements into an interface and the underlying algorithms falls within the analysis portion of the OOSE life cycle. The preliminary investigations determine the feasibility of system requested. There are three stages in the feasibility study portion of the preliminary investigation, namely, technical feasibility, operational feasibility and economical feasibility. Technical feasibility is concerned with specifying equipment and software that will successfully satisfy the user requirements and technical needs of the system. The operational feasibility deals with various operations on the software that should be easy and interactive. The economic feasibility deals with the tools required for developing and running the system, which should be easily available and cheap. In the present study, the selected tools are OSS which are free of cost and are downloaded from the web. The design phase involves the creation of geospatial health database. Spatial data have been created in the form of shape files. These shape files are used for retrieval, maintenance and deployment on the web. The thematic layers created include maps of India with State boundaries, Allahabad district with block boundaries, Malaria information for different blocks of Allahabad district. The parameters taken are Malaria susceptibility map, API, ABER and SPR for 2006, 2007 and 2008.

Health GIS 2009 register the user for authentication. After registration prototype Health SDI. The developed framework adopts process, user can use the framework with full phase a modular and flexible structure, and provides an operation. Module II gives detailed viewing of efficient mechanism for the generation and delivery of Allahabad district in terms of various factors associated value-added spatial information by extending the with Malaria disease and other data files. Module III concept of geospatial web services in the field of health describes utility services, i.e., management of user level sector. for security aspects, and uploading/ downloading features for full phase operation. The experience gained in using open source GIS software suggests that various tools and software like 5.5 Health SDI Model Observation Open JUMP, ALOV, Apache Tomcat, MySQL and PHP are available for creation of spatial datasets and The Open JUMP, ALOV, Apache Tomcat, MySQL, implementation of geospatial web services. The PHP and Java Development Tool Kit are used for widespread use of these open source resources in the overall development of the system. A part of Health SDI development of GIS based applications on the web framework thus developed has been shown using two instead of going for costly proprietary solutions could illustrations. Figure 5 shows State and Allahabad benefit a vast user community and should be District information using WMS. encouraged. Figure 5: Map showing States and Allahabad District REFERENCES Figure 6 shows information mapping for Malaria disease in different blocks of Allahabad district using WFS. Barik, R.K., Samaddar, A.B. and Gupta, R.D., 2009, In- vestigations into the Efficacy of Open Source GIS Figure 6: Information Mapping for Malaria using WFS Software, International Conference on Geospatial 6. CONCLUDING REMARKS Technology for Sustainable Planet Earth: Map World Forum- 09, Hyderabad, India. The present research work is focused at adopting OGC standards for creating, accessing, integrating and sharing Camara, G., 2004. Developing Open-Source GIS: What the geospatial health information on the web to develop are the Challenges? http://www.dpi.inpe.br/ GSDI, 2004. Developing Spatial Data Infrastructure: The SDI Cookbook, Ver. 2.0, http://www.gsdi.org/ docs2004/ Cookbook/cookbookV2.0.pdf Mall, R., 2004, Fundamentals of Software Engineering, 2nd Edition (India: Prentice-Hall) Mansourian, A., Rajabifard, A., Valadan Z.M.J. and Williamson, I., 2005, Using SDI and Web-based Sys- tem to Facilitate Disaster Management, International Journal of Computers & Geosciences, Vol. 32, pp.303-315 Puri, S.K., Sahay, S. and Georgiadou, Y., 2007, A Meta- phor-Based Sociotechnical Perspective on Spatial Data Infrastructure Implementations: Some Lessons from India, Research and Theory in Advancing Spa- tial Data Infrastructure Concepts, ESRI Press, pp.161-17 Rajabifard, A., Feeney, M.E.F. and Williamson, I.P., 2002, Future Directions for SDI Development, Inter- national Journal of Applied Earth Observation and Geoinformation, ITC, The Netherlands, Vol. 4, No.1, pp.11-22 Rawat, S., 2003, Interoperable Geo-Spatial Data Model in the Context of the Indian NSDI, Master's Thesis, ITC, The Netherlands Vanmeulebrouk, B., Rivett U., Ricketts, A. and Loudon M., 2008, Open source GIS for HIV/AIDS manage- ment, International Journal for Health Geographics,

Health GIS 2009 http://www.ij-healthgeographics.com/ content/ 7/ 1/ 53

HEALTH GIS FOR PART OF THIRUVALLUR TOWN USING CARTOSAT1 & LISS 4 PAN MERGED DATA S. Mohamed Ghouse*, J.Rathika**, V.Balamurugan, V.Ramakrishnan*** *Principal, ** Asst Professor, *** HOD Civil, MCA Sri Venkateswara college of Engineering Technology Tirupachur, Thiruvallur, Tamil Nadu, 631203 ABSTRACT: The 2.5m resolution PAN stereo data of Cartosat 1and 5.8 m Resolution Multi Spectral data Resourcesat Liss 4, of Indian Remote sensing satellites have been merged to create high resolution natural colour imagery of part of Thiruvallur town in Tamil nadu and it’s neighbourhood. The study area covers the segment covering North Western part of Thiruvallur town from the Temple Tank of SriVeeraragavasamy temple to junction of National Highways from Thiruvallur to Tirupathi and Thiruvallur- Poondi- Uthukottai state highway at the West, to create Health GISs to explain the use of High resolution Indian satellite data for local urban planning to manage the health and environmental issues. The ARC GIS and Leica photogrammetry software are used to create different layer of information on houses, roads, drainage, heath services like hospitals, pharmacy, emergency services like police, fire station, the wet lands, bushes and abandoned lakes which are the breeding grounds for mosquito’s in rainy season, surface drainage etc These layers are correlated with the municipal ward map of this segment of the town. It is observed that Liss 4, 5.8 m Resolution data is useful is creating GIS for health to understand problems from water bodies and abandoned lakes and cultivated lands in this town which is surrounded by cultivated paddy fields. The Cartosat 1 stereo data is useful for mapping the households, roads, agricultural fields, bushy areas and to understand the natural drainage of the area When the municipal ward maps are integrated with the GIS the drainage, drinking water lines, street names and house Nose etc can be added to the attribute data to make this as a complete Health GIS system. If 1m resolution Cartosat 2 data is used the individual house hold can be clearly mapped It is observed the PAN merged data of Cartosat 1 and Liss 4 meets the need of creating Health GIS as almost all information can be included in the GIS. It is also cost effective and suitable to understand the terrain condition due to the Stereo Capability of Cartosat 1. INTRODUCTION The segment covering the part of Thiruvallur town in Tamil Nadu from the Temple Tank of Sri Veeraragavasamy The high resolution Indian satellite data can be used for temple to West of National Highways from Thiruvallur to urban planning and creation of GIS for health, Tirupathi is selected for the study (Fig 1) to create Health environment, town planning, and drainage system. An GIS to demonstrate the use of high resolution Indian data for attempt has been made to demonstrate the use Indian local urban planning to understand the Health and satellites data using merged Cartosat 1, PAN 2.5m environmental problem areas to take remedial action by resolution stereo data dated 27th March 2007 and concerned. Resourcesat, Liss 4, 5.8 m, Resolution multi spectral data dated 11 Jan 2005 to create high resolution imagery of part The Arc GIs software is used to create different themes. of Thiruvallur town and it’s neighbourhood. The Leica photogrammetry GIS software is used to create the contours to understand the topographical features of the study area from the Cartosat 1 stereo data. The Resourcesat data of Liss 4 was useful in mapping with ease the agriculture lands and other natural resources like gardens, water bodies, urban areas, roads, paths etc as the data is in colour. In fact the merged data in true colour is very much useful in digitizing various layers to create the Health GIS. 1. AGRICULTURE LANDS & GARDENS Figure.1.Segment of the PAN &LISS 4 merged image of The study area consists of irrigated farm lands of study area paddy. The mango groves and tree plantations areas are also available in this area. These two are given in two layers in fig 2. The irrigated agriculture lands and grooves have stagnant water in rainy season and irrigated paddy season. These areas are breeding grounds of insects and mosquitoes etc

Agri lands & Gardens Thiruvallur NW drainage problem which causes stagnation of water in rainy season and creates health problems. The local N municipality is taking care to improve the roads and surface drainage and also laying underground drainage to WE solve the problem open drainage channels in the streets and health hazards. These problems are expected to be S solved when this project is completed in few years. Garden. Agri lands ThiruvallurNW- Houses ConcreteRoadsandEarthernRoads N W E S Figure 2. Agriculture & Garden lands And creates problems like fever and cold. The irrigated Building. agricultural lands also have reptiles like snakes which EarthenRd&Paths cause hazards like snake and insects bites. The terrain Highway&ConcreteRd condition which is undulating creates the surface drainage problems and stagnation of water etc these layers are Figure 4. Street houses and housing plots correlated with the municipal ward map of this segment of the town. 5. MARRIAGE HALLS, PUBLIC PLACES 2. URBANAISATION The small segment taken for study has more than 14 marriage halls, 6 hotels and 4 mosques, 6 schools, two The Town is growing in western and southern police station and part of Thiruvallur market. Apart from these direction after the location of the district collect orate. The town is urbanized at a faster rate. Hence urbanization is also an important factor to be considered while creating the health GIS. The fig 3 details the roads and highways. The study area is covered between the Thirutani – Thiruvallur national highway and Uttukkottai Thrivallure NE - High way & Concrete Roads,Contours – Thiruvallur state highway. Thrivallure NE - High way & Concrete Roads N E W N S WE S Contour 1m Building. High way &ConcreteRd Road.shp Figure 5 the Slope of study Area – Contour interval 1m Fig.3.Highways in Study Area The Veeraragavasmy temple and 7 temples are the places of social gathering for various functions. The health hazards due 4. BUILT UP AREA the public gathering during functions needs attention to keep the town clean from civic problems by removing the garbage and The old street houses, remodelled houses which are waste in time. The GIS of town will assist the concerned to get built closely are covered in this layer. These houses have rid of the health hazards caused by civic problems. the problem of ventilation and drainage. The layers of houses and the new housing plots area are added with the The GIS of the town will assist in furnishing the details road layer. The three layers are given in fig 4. New house to get rid of the health hazards caused by open drainage plots created by converting agriculture lands are also channels, public gathering in several places for festivals given as a separate layer. The new house plots are in low and functions. lying areas. The roads are only mud roads and there is

Thrivallore NE - Zones Ev1.shp N A T WE Od.shp Water.shp S Garden.shp Road.shp Water.shp Building.shp Study area.shp N Agricultural lan Garden WE Lay out New Urban Area 1 S New Urban Area 2 Town - Old Area Fig.6. Urban lands with buildings, roads, kutcha roads, tank Village UR 1 and water bodies in the study area. Village UR 2 6. The Satellite data with GIS layers Fig.8. Study Area Health Zones N Tiruvallur NorthWest- 8. CONCLUSIONS W E Cartosat +Liss4mergeddata  The GIS of the town can be studied by delineat- S ing in to Health Zones as detailed in Fig.8. 2  It is observed that Liss 4 5.8 m Resolution data is useful is creating GIS for Health to understand Fig.7. the GIS layers with satellite data problems from water bodies and abandoned The satellite data with all layers are given above. The lakes and cultivated lands for this town which image 7 indicates the low lying areas, the old town and the this town is surrounded by cultivated paddy housing plots area. The paddy fields with crop and without fields. crops and gardens can also be seen.  The Cartosat 1 stereo data is useful for mapping 7. ZONATION TO SOLE HEALTH ISSUES the households, roads, agricultural fields, bushy Based on this the segment can be delineated in to areas and to understand the natural drainage of zones as detailed in Fig 8 to solve the health and the area. environmental issues. The zones are agriculture lands, Villages, old town, housing plots, layouts, new housing  When the municipal ward maps are integrated colonies etc. This grouping will make the concerned with the GIS the drainage, drinking water lines, authorities to draw a plan to solve the health issues in the street names and house No’s etc can be added to study area. the attribute data.  If Cartosat 2 1m data is used the individual house hold can be clearly mapped. It is observed that the PAN merged data of Cartosat 1 and Liss 4 still meets the need of Health GIS as almost all information can be given and It will be cost ef- fective and most suitable to understand the ter- rain due to the Stereo Capability of Cartosat 1.

BUILDING A GEOREFERENCED POPULATION HEALTH REGISTRY Seetharaman Narayanan, Chandrasekaran Ramaswamy , Christopher Amalraj Vallabadas, Hari Srimathi, Shanmuganathan Pasupathy, Vadivu Ganesan Department of Community Medicine, SRM Medical College & Research Centre, SRM University, Department of Community Medicine, SRM Medical College & Research Centre, SRM University, Information Technology, Faculty of Engineering & Technology, SRM University, Computer Applications, Faculty of Engineering & Technology, SRM University, Civil Engineering, Faculty of Engineering & Technology, SRM University, Department of Community Medicine, SRM Medical College & Research Centre, SRM University. ABSTRACT: In the field of Public Health, Geographical Information Systems (GIS) serve as a potent tool in planning, analysis, monitoring and management of health systems. GIS is also a promising tool for multi-disease surveillance. Despite the tremendous potential of GIS, the health sector in India has not fully utilized it. The Population Health Registry (PHR) project in the Kattankulathur administrative block near Chennai, Tamilnadu proposes to cover approximately 150,000 people living in over 65 villages. Initial data collection for registry building would be for over a period of three years and has both cross-sectional surveys and periodical updates utilizing local talent. The data collected has two distinct sources – a) GIS based mapping and b) house-to-house surveys. The GIS database contains records of individual households, water sources, health facilities and other socio- environmental variables related to health. Field Surveys would capture key demographic and health data at three levels – the village, household and individual levels. Unique Identifiers would be assigned to every individual living in the 65 villages. The PHR would serve as an excellent platform for visualizing and analyzing epidemiological data, for revealing trends, dependencies and inter-relationships between health and socio-environmental factors. The registry is NOT a Surveillance system in the true sense, but uses data from a micro-level surveillance using local talent. A unique query based Health Management Information System (HMIS) would be built by linking the SRM Hospitals’ medical records with the georeferenced PHR Initially the data collected pertaining to diseases would be limited to self-reported illnesses. Further ‘layers’ of health data collection, with clinical and possible bio-chemical examinations and disease specific add-on modules are planned at later stages once the basic data-capturing mechanism is in place and the registry has the necessary baseline data. Thus the registry would provide valuable baseline data for numerous community based health and social sciences research. Geospatial representation of health and related environmental data, on this scale, has not been earlier attempted in India. This will also compliment the Government’s ongoing Integrated Disease Surveillance Programme. This project is being is undertaken in technical collaboration with Queen’s University Ontario, Canada through an MoU KEY WORDS: GIS, Health, PHR, Population Health Registry, India. 1. INTRODUCTION To develop effective interventions for reducing the morbidity and mortality from many diseases, we need The use of Geographical Information Systems (GIS) in access to micro-level demographic, environmental and Health research has been rapidly increasing. Instances of epidemiological data. The science of Epidemiological application of GIS technology in the health field include Intelligence and systems of Disease Surveillance are still disease surveillance and epidemic forecasting, health- rudimentary in India. The lack of regional capacity to services availability mapping, vector mapping, better collect and process health data adds to the problem. The trauma care and much more. In the field of Public Health, PHR aims to empower regional capacity for health data GIS has opened up exciting new opportunities for health collection and analysis through a Participatory approach administrators to enhance planning, analysis, monitoring in building the Registry. and management of health systems. Spatial Epidemiology has acquired a whole new dimension with GIS providing A Registry is an organized system for the collection, an excellent means of visualizing and analyzing storage, retrieval, analysis, and dissemination of epidemiological data, for revealing trends, dependencies information on individual persons who usually have a and inter-relationships between health status and socio- particular disease or exposure of interest. Disease-specific environmental variables. Despite the tremendous registries – like cancer registries - are being maintained in potential of GIS, the health sector in India has not fully India and the world over, usually by Governments and utilized its capabilities. Universities. However a comprehensive health registry that captures the entire spectrum of socio-demographic

and environmental factors in the form of a Community Figure1 Pilot villages in the Kattankolathur Block Health Profile is yet to be built. 2.3 Data Collected Data collection involves two distinct components – the The proposed Population Health Registry (PHR) is GIS based mapping and the house-to-house survey unique in that it integrates individual health records with component georeferenced household and socio-environmental 2.3.1 GIS aided mapping variables. The PHR would be a web-based GIS system - Individual households, water sources, health facilities and Health data is stored in a central server which can be environmental variables that have a bearing on the health accessed from various terminals connected to the server status of the population would be mapped and geo- through internet or intranet. Internet based GIS referenced. (See Annexure I for the complete list of study technology eliminates the traditional method of flow of variables). Most of the GPS data are one-time information, and the information is instantly available measurements, while a select few variables need to be across the globe periodically updated (migration, new constructions etc). 1.1 Objectives Figure 2 An overlay of GPS coordinates of village data over Google Earth© Satellite Imagery 1. To establish a population health registry (PHR) through census enumeration, capturing key health-related 2.3.2 Field Surveys environmental and socio-demographic variables in 65 For capturing key health and relevant demographic & villages in Kattankulathur Block socio-environmental data would be carried out at three 2. To digitally map using the Geographical Information System (GIS), the eco-social factors related to health and disease in these villages 3. To link the PHR census data to the geo-spatial database acquired through GIS mapping thereby creating a query- based Health Management Information System (HMIS) 4. To link the medical records of SRM MCH and its outreach centres with the PHR, creating avenues for future research 2. METHODOLOGY 2.1 Scope of the Registry The PHR is not a disease-specific registry. It is a population health registry – it contains socio- demographic and environmental data related to the health status of the population. Most data, including health & medical information is self-reported and there are no clinical or laboratory examination involved 2.2 Population covered The registry covers over 150,000 people living in 65 villages of the Kattankolathur administrative block near Chennai, Tamilnadu. 60 villages that form the catchment area of the SRM Hospitals would form the study area. 5 villages are currently under the pilot phase. Within each village, since every individual would be covered through census enumeration, no specific sampling technique needs to be used. Each individual would be assigned a ‘unique identifier’

levels – the village, the household and the individual 3. RESULTS AND DISCUSSION levels. (See Annexure II for the list of Questionnaires) This is essentially the ground report of the pilot phase of Unique identifier numbers are given to every individual a unique multi-disciplinary effort at building a geo- in every village. At the individual level, different sets of referenced population health registry. The lessons we questionnaire are employed for special groups – children, learnt from piloting may be shared with the wider antenatal mothers, women in reproductive age group and research community. The research team includes the elderly. Overall, we are studying 650 variables that are Departments of Community Medicine, Civil Engineering, directly or indirectly related to the health status of the Information Technology, Microbiology, Computer individual and the community Science and the Schools of Public Health and Biotechnology 2.4 Data Entry, Storage & Processing The PHR gives us access to georeferenced health data at Oracle 10 G and Visual Studio 2005 has been used in the family and even individual levels. Thus the registry designing and developing the database that forms the would serve as fertile grounds for exploring and registry. The user interface (data input screen) has been establishing correlations between diseases / conditions designed with local language support and has extensive and exposure to various environmental variables or to use of click and drop-down menus. track the spread of the disease temporally 2.4.1 Geospatial Data The PHR provides us a snapshot of health in the form of a community Health Profile - a set of health, GPS data collection is done using a Garmin GPSMAP demographic and socioeconomic indicators which are 60CS and a Tremble Juno SB handheld. ESRI ArcGIS relevant to most communities. This also enables Desktop 9.2 has been used for handling geospatial data. comparisons locally, regionally and over time. A team of three GIS mappers map a village in ten working days. 3.1 Issues of Interest 2.4.2 Survey Data Scarcity of suitable local talent: There have been high drop-out rates among village volunteers after completing Local youth (two per village, one female and one male) the hands-on training on usage of laptops and data entry. have been trained as Community Health Surveyors for PDAs: With their ease-of-use and compactness, PDAs survey data collection. Survey tools and techniques have have evoked better responses from both the surveyors and been standardized. Data collection and data entry is done, the surveyed. as one step, directly on to laptops with local language Add-on Modules: The first layer of Health data collected support. Personal Digital Assistants (PDAs) are also during the survey is primarily from self-reported illness. being tested for data collection. Necessary Quality checks Once the basic data collection mechanism is in place, have been incorporated and the survey data is transferred further layers of health data can be gathered in the form to the secure servers at the university. of add-on modules. Currently an add-on module on the nasal carriage of Community Acquired Methicillin- 2.5 The Health Registry resistant Staphylococcus Aureus (CA-MRSA) is being carried out. The registry would combine the GIS database and the PGIS: By involving the local community in GIS data survey data to produce a unique query-based Health gathering, a Participatory GIS (PGIS) approach is Management Information System (HMIS). Periodic envisaged. updation of selected variables is built in to the system. Google Earth© overlays have a great visual appeal among The system can be queried to display, for instance – all the villagers houses in the village > houses of all known diabetics > houses of all known female diabetics above 50 yrs > 4. CONCLUSIONS houses of all known female diabetics above 50 yrs who have cattle in house / use a particular source of drinking The PHR provides access to geo-referenced micro- water – a choice of endless possibilities. level demographic, environmental and epidemiological data with unprecedented levels of detail. The registry is 2.6 Linkage with Hospital Records fertile grounds for numerous health and social sciences research. Involving local talent in registry building is a Linking with medical records from SRM hospitals and right step towards empowering regional capacity for its two outreach centres would be done once the initial health data collection and analysis. This also seeds the stages of registry-building are completed and registry has community with the science of Epidemiological sufficient baseline data. Individual patients may be Intelligence at the grass-roots level. tracked over time and their relation to environmental variables explored