GISdevelopment.net --> Application --> Health
GIS to the Rescue – Controlling the Killer AIDS
Sudeep.K.V
Head – Marketing
IES Geospatial Systems India PVT LTD.
Email: sudeep@iesgeospatial.com
Sudeep.K.V
Head – Marketing
IES Geospatial Systems India PVT LTD.
Email: sudeep@iesgeospatial.com
Introduction
This paper presents the use of GIS in controlling and monitoring the spread of AIDS by providing information well in advance on the potential budding high transmission sites. It also helps the authorities in the planning process by identifying suitable sites to set up facilities required in the fight against AIDS. The central idea is to make the fight against AIDS pro-active rather than reactive.
GIS and Mapping technology are used to graphically represent the data and derive relations between the transmission sites and other parameters. This paper attempts to explore the powers of simple thematic mapping and map query. Thematic mapping and query when combined with simple functions offered by most GIS software become an unmatched combination of technology to provide valuable solutions to the complex problems faced by health authorities today.
This very technology is used in this paper to analyze and thematically map HIV/AIDS data of transmission areas from 14 provinces within Kerala in correspondence with their local population statistics to cartographically portray the demographic consequences that HIV/AIDS would evoke at a provincial level. We begin from a global perspective and focus down to specific area and refine our finding and provide recommendations. To keep the paper concise and tight only 2 transmitter categories have been analyzed in detail with relation to the demographics data available.
Methodology
The process of identifying potential high transmission sites begins by analyzing the complete state. There are many data requirements to enable the functioning of the GIS system. Kerala base administrative layers for districts and taluks are digitized. High transmission sites already identified have been collected from the AIDS control society. Demographic data has been made available by census department of India. All other major towns, cities and tourist places have been digitized form survey of India sheets.
Once all the layers are made available in digitized form the data is attached to the map. In this case the data resides in Microsoft Excel format and the same is linked to the GIS system by ID matching process.
The 1st step is to identify all the risk transmitters’ category and like wise all the factors that propagate the growth of these categories. This is mostly a manual effort that becomes apparent from working in the field. The AIDS control society of India has identified the following as high risk transmitters…
We begin from a global perspective with the smallest unit being a taluk. The initial parameters for the study is to analyze for demographics Vs the spread of infected CSW across the state.
Fig.1 is the thematic map of taluk representing the spread of infected CSW. The intervals chosen are 4 in this case. At this stage we are analyzing from a global perspective for the purpose of focusing on specific areas. From the thematic map it is apparent that there are clearly segregated areas with high presence of infected CSW. For a realistic approach we confine the analysis to the top 3 interval areas of the thematic map.
The next step begins with selectively querying for population. Population query is generated for both urban as well as rural areas. In each case the query begins with the maximum population figure. We start to reduce the query figures to see the results graphically. With incremental reduction in the base figure the area selected will increase. The idea is to use the population query (Greater than operator) on a trial and error basis till at least 75% of the areas that represent high on infected CSW count is intersected by the query. This is represented in Fig. 2. During this process there will be a couple of areas that are currently not in the high risk areas getting selected. These areas become potential sites for further analysis.
The areas selected as a result of the above analysis is represented in Fig. 3. At this stage it may be noted that many of the sites may fall off from the potential site list as we begin to analyze for other parameters.
The next phase of the analysis will still be done on the infected CSW transmitters but this time we use the rural population for query.
The process for query remains the same. We use trial and error method to cover at least 75% of the taluks with high infected CSW population. While doing this it becomes evident that inspite of reducing the rural population figures substantially we fail to cover 75% of the taluks infested with infected CSW. Is this an indication that infected CSW are less in rural areas?
Strikingly we see that 2 taluks in Trivandrum district qualify relatively early while querying for rural population. The reason for this is not analyzed in this paper. From this query result we arrive with a general conclusion that areas with high rural population is generally free from infected CSW. With the exception of 2 taluks represented in Fig.4. Till this point we have seen that there is a direct proportion between infected CSW and urban population. It has also been concluded that the spread in rural areas is less.
Now we analyze for relations between infected CSW and other categories to look for possible trends. To begin with we look at the spread of infected CSW Vs infected truckers. Here we use a simple technique. Infected CSW and truckers spread is queried for number of cases till 50% of the study area has been selected (Kerala in this case). In the next step we intersect both the selections. An overlap of 70% or greater is a clear indication that there is a direct proportional relationship between infected CSW and truckers. This is represented in Fig.5
Fig.5a represents the query selection of infected CSW. Areas selected are those having infected CSW population greater than 50. There are 28 areas selected in this query. Fig.5b represents the infected truckers spread. The number of records selected in this case is 35. On intersecting it is found that 21 areas get selected which is an indication that there exists a relation between the truckers and CSW spread. The CSW population is prone to migration in many cases. They tend to move to more desirable areas from their perspective. In Fig.5c we isolate all areas that are high in truckers spread but are relatively very low on the CSW count. The number of such records as seen from Fig.5c is 14. These are potential sites for migration of CSW. At this stage we further consolidate our analysis by intersecting the results of the above analysis with the analysis of infected CSW spread based on rural and urban population. In this exercise we are still analyzing from a global perspective to focus on specific regions. The results of the above analysis are graphically illustrated in Fig.6. There are 2 taluks that have been selected as having potential to harbour CSW. We now take a look at the figures that have been collected by the AIDS control society. The figures are shown below. It is clearly evident that the analysis is accurate enough in isolating high transmission areas.
The 2 areas isolated were Aluva in Cochin and Mukundapuram in Trichur district. The current figures of infected CSW in these 2 areas are…
At this stage we complete the global analysis and confine our analysis to the areas that are isolated. Now we would be confining our study to the 2 talluks, Aluva and Mukundapuram.
In Fig.7a and 7b we thematically map the high transmission site for 2 parameters i.e for infected CSW and truckers. A look at the map reveals an offset or imbalance between the CSW infected and the truckers infected in 2 sites. Namely Kondallor and Kalladi.
These are potential sites for migration to take place. Further analysis is done by creating buffers on all the major roads to see how the transmission sites are located with respect to the road. In this case we create 1 Km buffer on all the major roads and find that all high transmission sites are within this range.
After the buffer creation we switch on the layers representing all major locations that have at least any one of the following…
Fig. 8
While creating buffers couple of places (represented by point theme) intersect with the road buffers. We retain only such points for further analysis. This is represented in Fig.9. PHS specializing in monitoring AIDS either through NGO or by government initiative has been represented by switching on the relevant layer. This is represented in Yellow.
Fig. 9
All the points that remain now are the areas that have to be watched for transmission through infected CSW and truckers. The agencies that are entrusted with this exercise are spread across the taluk. It has been confirmed by research that PHC are effective in 5Km radius of their operation. Areas beyond 5Km are normally neglected and if at all any activity takes place it is only reactive in nature.
To analyze the spread of the PHCs we create buffers of 5Km radius to each PHC to see the areas that are being effectively covered. This is represented in Fig.10. It is seen that there are 7 potential sites that can harbor AIDS carriers and which do not fall under the preview of any PHC. Strikingly there are 3 already high transmission sites that also do not fall in the effective preview of any PHC. There is one more point to be noted that there are a lot of areas that fall under the preview of more than 1 PHC (Overlaps).
Fig. 10
Inspite of having 11 PHC dedicated to the monitoring of AIDS their placement have not been rationalized. The next figure represents the rationalization of these PHC.
Fig. 11
We have still retained the same number of PHCs but have covered all the potential sites. This has been done by moving the buffer areas to cover all the spots. In the event of all the spots not getting covered it calls for new PHCs to be set up.
In this analysis we have used the demographics, road networks and select sites. Land use, terrain and other spatial data can also be used to make more accurate analysis. In this illustration we arrive at 2 taluks for the final analysis. By selectively varying the base query figures more areas can be selected in the order of their sensitiveness to be AIDS prone. We could then analyze all these areas on various parameters using the same technique as illustrated.
Conclusion
To summarize we have used thematic mapping, query and buffer operations selectively for drawing general conclusion from the global perspective and then focusing down to selective areas for further analysis. We have identified the sites that are prone to harbor select category of transmitters. We have further analyzed the PHC spread and checked them for their location advantages. We have also reallocated these PHCs for optimization. In this paper we have completed the entire process of analysis from basic analysis to making recommendation for relocations and pin pointing potential high transmission sites. The idea was to explain the process rather than to isolate all the potential high transmission places in the entire state.
As you have seen this paper has demonstrated the use of simple thematic mapping, querying and a few other features to actually strategize the whole fight against AIDS by bringing to light all vital spatial information and relations. Is the day near when we could see GIS being used extensively in India to the fight against the killer AIDS?
This paper presents the use of GIS in controlling and monitoring the spread of AIDS by providing information well in advance on the potential budding high transmission sites. It also helps the authorities in the planning process by identifying suitable sites to set up facilities required in the fight against AIDS. The central idea is to make the fight against AIDS pro-active rather than reactive.
GIS and Mapping technology are used to graphically represent the data and derive relations between the transmission sites and other parameters. This paper attempts to explore the powers of simple thematic mapping and map query. Thematic mapping and query when combined with simple functions offered by most GIS software become an unmatched combination of technology to provide valuable solutions to the complex problems faced by health authorities today.
This very technology is used in this paper to analyze and thematically map HIV/AIDS data of transmission areas from 14 provinces within Kerala in correspondence with their local population statistics to cartographically portray the demographic consequences that HIV/AIDS would evoke at a provincial level. We begin from a global perspective and focus down to specific area and refine our finding and provide recommendations. To keep the paper concise and tight only 2 transmitter categories have been analyzed in detail with relation to the demographics data available.
Methodology
The process of identifying potential high transmission sites begins by analyzing the complete state. There are many data requirements to enable the functioning of the GIS system. Kerala base administrative layers for districts and taluks are digitized. High transmission sites already identified have been collected from the AIDS control society. Demographic data has been made available by census department of India. All other major towns, cities and tourist places have been digitized form survey of India sheets.
Once all the layers are made available in digitized form the data is attached to the map. In this case the data resides in Microsoft Excel format and the same is linked to the GIS system by ID matching process.
The 1st step is to identify all the risk transmitters’ category and like wise all the factors that propagate the growth of these categories. This is mostly a manual effort that becomes apparent from working in the field. The AIDS control society of India has identified the following as high risk transmitters…
- Commercial sex workers (CSM)
- Male sex male (MSM)
- Intravenous drug users (IVDU)
- Fishermen
- Migrants
- Construction workers
- Street Children
- Truckers
- Population (Rural and Urban)
- Proximity to major road networks
- Proximity to railway network
- Basic type of place (Tourist / business etc.)
We begin from a global perspective with the smallest unit being a taluk. The initial parameters for the study is to analyze for demographics Vs the spread of infected CSW across the state.
Fig.1 is the thematic map of taluk representing the spread of infected CSW. The intervals chosen are 4 in this case. At this stage we are analyzing from a global perspective for the purpose of focusing on specific areas. From the thematic map it is apparent that there are clearly segregated areas with high presence of infected CSW. For a realistic approach we confine the analysis to the top 3 interval areas of the thematic map.
The next step begins with selectively querying for population. Population query is generated for both urban as well as rural areas. In each case the query begins with the maximum population figure. We start to reduce the query figures to see the results graphically. With incremental reduction in the base figure the area selected will increase. The idea is to use the population query (Greater than operator) on a trial and error basis till at least 75% of the areas that represent high on infected CSW count is intersected by the query. This is represented in Fig. 2. During this process there will be a couple of areas that are currently not in the high risk areas getting selected. These areas become potential sites for further analysis.
The areas selected as a result of the above analysis is represented in Fig. 3. At this stage it may be noted that many of the sites may fall off from the potential site list as we begin to analyze for other parameters.
The next phase of the analysis will still be done on the infected CSW transmitters but this time we use the rural population for query.
The process for query remains the same. We use trial and error method to cover at least 75% of the taluks with high infected CSW population. While doing this it becomes evident that inspite of reducing the rural population figures substantially we fail to cover 75% of the taluks infested with infected CSW. Is this an indication that infected CSW are less in rural areas?
Strikingly we see that 2 taluks in Trivandrum district qualify relatively early while querying for rural population. The reason for this is not analyzed in this paper. From this query result we arrive with a general conclusion that areas with high rural population is generally free from infected CSW. With the exception of 2 taluks represented in Fig.4. Till this point we have seen that there is a direct proportion between infected CSW and urban population. It has also been concluded that the spread in rural areas is less.
Now we analyze for relations between infected CSW and other categories to look for possible trends. To begin with we look at the spread of infected CSW Vs infected truckers. Here we use a simple technique. Infected CSW and truckers spread is queried for number of cases till 50% of the study area has been selected (Kerala in this case). In the next step we intersect both the selections. An overlap of 70% or greater is a clear indication that there is a direct proportional relationship between infected CSW and truckers. This is represented in Fig.5
Fig.5a represents the query selection of infected CSW. Areas selected are those having infected CSW population greater than 50. There are 28 areas selected in this query. Fig.5b represents the infected truckers spread. The number of records selected in this case is 35. On intersecting it is found that 21 areas get selected which is an indication that there exists a relation between the truckers and CSW spread. The CSW population is prone to migration in many cases. They tend to move to more desirable areas from their perspective. In Fig.5c we isolate all areas that are high in truckers spread but are relatively very low on the CSW count. The number of such records as seen from Fig.5c is 14. These are potential sites for migration of CSW. At this stage we further consolidate our analysis by intersecting the results of the above analysis with the analysis of infected CSW spread based on rural and urban population. In this exercise we are still analyzing from a global perspective to focus on specific regions. The results of the above analysis are graphically illustrated in Fig.6. There are 2 taluks that have been selected as having potential to harbour CSW. We now take a look at the figures that have been collected by the AIDS control society. The figures are shown below. It is clearly evident that the analysis is accurate enough in isolating high transmission areas.
The 2 areas isolated were Aluva in Cochin and Mukundapuram in Trichur district. The current figures of infected CSW in these 2 areas are…
-
Aluva : 39 Cases
Mukundapuram : 48 Cases
At this stage we complete the global analysis and confine our analysis to the areas that are isolated. Now we would be confining our study to the 2 talluks, Aluva and Mukundapuram.
In Fig.7a and 7b we thematically map the high transmission site for 2 parameters i.e for infected CSW and truckers. A look at the map reveals an offset or imbalance between the CSW infected and the truckers infected in 2 sites. Namely Kondallor and Kalladi.
These are potential sites for migration to take place. Further analysis is done by creating buffers on all the major roads to see how the transmission sites are located with respect to the road. In this case we create 1 Km buffer on all the major roads and find that all high transmission sites are within this range.
After the buffer creation we switch on the layers representing all major locations that have at least any one of the following…
- Parks
- Bus station
- Tourist attraction
- Business centers
Fig. 8
While creating buffers couple of places (represented by point theme) intersect with the road buffers. We retain only such points for further analysis. This is represented in Fig.9. PHS specializing in monitoring AIDS either through NGO or by government initiative has been represented by switching on the relevant layer. This is represented in Yellow.
Fig. 9
All the points that remain now are the areas that have to be watched for transmission through infected CSW and truckers. The agencies that are entrusted with this exercise are spread across the taluk. It has been confirmed by research that PHC are effective in 5Km radius of their operation. Areas beyond 5Km are normally neglected and if at all any activity takes place it is only reactive in nature.
To analyze the spread of the PHCs we create buffers of 5Km radius to each PHC to see the areas that are being effectively covered. This is represented in Fig.10. It is seen that there are 7 potential sites that can harbor AIDS carriers and which do not fall under the preview of any PHC. Strikingly there are 3 already high transmission sites that also do not fall in the effective preview of any PHC. There is one more point to be noted that there are a lot of areas that fall under the preview of more than 1 PHC (Overlaps).
Fig. 10
Inspite of having 11 PHC dedicated to the monitoring of AIDS their placement have not been rationalized. The next figure represents the rationalization of these PHC.
Fig. 11
We have still retained the same number of PHCs but have covered all the potential sites. This has been done by moving the buffer areas to cover all the spots. In the event of all the spots not getting covered it calls for new PHCs to be set up.
In this analysis we have used the demographics, road networks and select sites. Land use, terrain and other spatial data can also be used to make more accurate analysis. In this illustration we arrive at 2 taluks for the final analysis. By selectively varying the base query figures more areas can be selected in the order of their sensitiveness to be AIDS prone. We could then analyze all these areas on various parameters using the same technique as illustrated.
Conclusion
To summarize we have used thematic mapping, query and buffer operations selectively for drawing general conclusion from the global perspective and then focusing down to selective areas for further analysis. We have identified the sites that are prone to harbor select category of transmitters. We have further analyzed the PHC spread and checked them for their location advantages. We have also reallocated these PHCs for optimization. In this paper we have completed the entire process of analysis from basic analysis to making recommendation for relocations and pin pointing potential high transmission sites. The idea was to explain the process rather than to isolate all the potential high transmission places in the entire state.
As you have seen this paper has demonstrated the use of simple thematic mapping, querying and a few other features to actually strategize the whole fight against AIDS by bringing to light all vital spatial information and relations. Is the day near when we could see GIS being used extensively in India to the fight against the killer AIDS?