Toolkit_MAP_for_Access_Studies
Contents
1. psi Healthy lives Measurable results PSI Research amp Metrics Toolkit Building Research Capacity MAP FOR ACCESS STUDIES DESIGN IMPLEMENTATION AND UTILIZATION PSI s Four Pillars Bottom Line Health Impact Private Sector Speed and Efficiency Decentralization Innovation and Entrepreneurship Long term Commitment to the People We Serve Research amp Metrics Population Services International 1120 Nineteenth Street NW Suite 600 Washington D C 20036 MAP for Access Studies Design Implementation and Utilization PSI Research amp Metrics 2011 Population Services International 2011 Contacts Bram Piot MAP amp GIS Coordinator Population Services International 1120 19 Street NW Suite 600 Washington DC 20036 P 202 785 0072 E bram piot orange fr Virgile Capo Chichi Senior Technical Advisor Population Services International 1120 19 Street NW Suite 600 Washington DC 20036 P 202 785 0072 E vcapo chichi psi org LEARNING OBJECTIVES By the end of this chapter the reader will be able to Describe the objectives and characteristics of Measuring Access for Performance MAP for access surveys Understand the advantages and limitations of access surveys Design and implement an access survey and prepare data collection instruments Prepare data for analysis as well as analyze report and interpret data BACKGROUND This R amp M Toolkit chapter provides guidance to researchers wh2. Jem Lo T a k ak 7 a com GoOgle E E l Imagery Date 1211 2008 2p 2009 32B lend Eyeall 187km r Most importantly the summary report should include a number of programmatic recommendations based on the study findings These should be discussed with the programmers and marketers soon after the analysis is completed and before the summary report is finalized This discussion should take place during a formal Dashboard to Decision making DDM session DDM session outputs are a series of programmatic recommendations and an action plan that addresses coverage gaps and issues with quality of coverage These outputs should be summarized in the Programmatic Recommendations section of the MAP report A separate MAP DDM guidance document describes in detail the entire process and outputs of a DDM session 22 SUMMARY OF IMPLEMENTATION STEPS FOR A MAP FOR ACCESS STUDY Step One Define MAP Study Objectives Step Two Define the Unit of Analysis and Minimum Standards for Access Step Three Define Appropriate Stratification Step Four Benchmarking for equity of access if relevant Step Five Create the Sampling Frame Step Six Determine the Sampling Procedure Step Seven Design the Audit Sheet Step Eight Define Data Collection Procedures Step Nine Establish Timeline and Budget Step Ten Set Up GIS Step Eleven Pilot Test the Audit Sheet and Data Collection Procedures Step Twelve Train Researchers Interviewers and Supervisors
3. Step 6 Compute the access rate To do so divide the number of people inside the catchment area those with access by the total number of people in the target group This last step is usually done in Excel or manually with a calculator based on the numerator and denominator obtained in the GIS Note that the general procedure described above may differ from one software package to another Even within the same software there may be several possible combinations of functions and tools to obtain the correct result All of these variations and tools are covered in detail during the GIS training Appendix 2 explains one of the methods that can be used in ArcGIS to calculate access See Figure 6 for an illustration of the different types of data layers and how these are combined in order to enable access analysis in a GIS Figure Six Calculating Access in a GIS Required Information Layers Hotzone administrative boundaries Type of target group Target group distribution Catchment area buffer Product Service detivery point locations 19 STEP 15 REPORT WRITING AND DASHBOARD TO DECISION MAKING SESSIONS The standard MAP Summary Report template should be followed to present the findings of the Access study It should also contain a number of relevant programmatic recommendations based on these findings in view of improving or maintaining adequate access levels The report template contains instructions on the format and conte
4. who are dressed like women Hotzones areas defined by clustering hotspots with contiguous catchments 25 Nepal Hotzone a cluster of points with contiguous catchments where the high risk target group is active For FSWs these points include streets where sex workers operate truck bus stops bars cabin restaurants dance restaurants or hotels The outer perimeter of a hotzone will ultimately be defined by programmatic need and may vary between urban and rural areas Papua New Guinea Hotzone area where there is a concentration of establishments hotspots in which members of the target group are present They may correspond to a neighborhood or a section of a street Census units are suitable for measuring the availability of condoms in the provincial capitals and in the selected enclaves whereas hotzones are more appropriate when measuring access to condoms among specific at risk groups such as sex workers Zimbabwe Hotzone a catchment area within a radius of 1km of a condom outlet in rural areas and within 500m in high risk areas in urban sites 26 Appendix 2 Summary Steps for Calculating Access in ArcGIS version 9 Step 1 Open ArcMap and display the two layers that you need to calculate access 1 the layer with population data e g number of clients per hotspot and 2 the layer with product availability e g number of condoms by outlet Use a shapefile extracted from the HealthMapper Data Temp folder or ad
5. Piot B 2011 PSI Google Mapping Toolkit PSI Research Toolkit Series Washington DC Population Services International Available at http www psi org resources publications Grossman LS 1986 Consumer behavior and the village trade store a Papua New Guinea example Geografiska Annaler Series B Human Geography 68 1 41 49 Noor AM et al 2006 Modelling distances travelled to government health services in Kenya Tropical Medicine and International Health 11 2 188 196 Piot B amp Capo Chichi V 2011 MAP for coverage studies design implementation and utilization PSI Research Toolkit Series Washington DC Population Services International Available at htto www psi org resources publications Piot B Raharinjatovo J White J Hobby N Rabemanatsoa A Morgan G amp Patel D 2006 Tools PDAs for data collection PSI Research Toolkit Series Washington DC Population Services International Rosero Bixby L 2004 Spatial access to health care in Costa Rica and its equity a GlS based study Social Science and Medicine 58 1271 1284 24 Appendix 1 Sample Hotzone Definitions from Various PSI Platforms The following definitions of high risk areas hotzones or hotspots for HIV AIDS are based on a non exhaustive review of MAP study designs These definitions differ from platform to platform and are given as examples only Burundi Hotzone an area within a 500m radius of a hotspot India Category A hots
6. Step Thirteen Conduct Auditing during Data Collection Step Fourteen Analyze Data Step Fifteen Hold DDM and Report on Findings QUALITY IMPROVEMENT CHECKLIST To ensure the MAP for Access study is effective and produces high quality outputs use the following checklist The objectives of the study are clearly defined and understood by the research and program teams The study objectives and indicator definitions are discussed with program managers A study design that clearly describes the primary sampling unit geographic unit of analysis the sampling frame and indicator definitions minimum standards exists Data collection tools are designed to collect product or service availability The audit sheet s and data collection procedures are pretested All interviewers and supervisors are trained on data collection procedures including use of GPS units Researchers have the capacity to analyze access Analysis is performed within a few weeks of data entry data upload A DDM session is conducted with the program teams shortly after analysis is completed and before the report is written The report follows the standard MAP Summary Report template 23 REFERENCES Chapman S Capo Chichi V Longfield K amp Piot B 2005 Project MAP measuring access and performance Lessons learned and recommendations Washington DC Population Services International Available at http www psi org resources publications Duvall S and
7. collectors record these with GPS devices or with GPS enabled Personal Digital Assistants PDAs see the next step for information on recording format and procedures In addition to the coordinates the audit form should also include clear identification information for each site Typically this identifying data includes the name and or code of the administrative area region province district the municipality or village the neighborhood or hotzone and the street address if relevant 13 2 Type and size of target group This information is key for calculating access since access levels cannot be determined without information on the number of people that work meet entertain reside in each of the sites Usually the target group refers to the hotspot audience such as the approximate daily number of clients in a bar the number of FSWs in a brothel the number of MSM in a night club etc Collect audience data only in the target audience sites not in the outlets or service providers While it is usually not feasible to accurately determine the number of members of the target group of interest because of time funding constraints it is generally possible to obtain an estimate on the number of clients workers residents from a key informant such as the bartender night club manager or another person who is familiar with the venue Because client numbers at these venues may vary from day to day e g weekend vs day in the week as well as throug
8. e he OFRP PrP OF RP FP OF O When calculating this simplified access rate the underlying assumption is that all individuals who live in the community are considered having physical access to the provider Note that this method results in realistic access estimates only if the size of the geographic unit of analysis for which population data are available roughly corresponds to the size of the catchment area of the product or service In such studies access rates are generally similar to average coverage measures though the simplified access definition will adjust for variations in population size of the sampled areas This approach mainly applies to studies on access to products or services that relate the general population or to population segments that are not highly concentrated in geographic locations e g malaria risk strata Equity of Access Equity of access is defined as equal levels of access to product or service delivery points at the minimum access standard among population segments with equal levels of need risk or demand Inequity of access then is when population segments do not have equal levels of access PSI measures the extent of inequity of access as the difference in access between the population with the highest level of access and other populations with similar levels of need For example among high risk populations for HIV AIDS access to condoms among FSWs men who have sex with men MSM and truck driv
9. section describes the steps to follow when designing and implementing an access study Since the majority of access studies also measure the coverage and quality of coverage indicators the reader should refer to the Coverage toolkit for general guidance on how to define minimum standards and supervision areas on sampling and on benchmarking The same steps are followed here This toolkit reviews in detail those aspects that are specific to access studies IMPLEMENTATION STEPS FOR A MAP FOR ACCESS STUDY Step One Step Two Step Three Step Four Step Five Step Six Step Seven Step Eight Step Nine Step Ten Step Eleven Step Twelve Step Thirteen Step Fourteen Step Fifteen Define MAP Study Objectives Define the Unit of Analysis and Minimum Standards for Access Define Appropriate Stratification Benchmarking for equity of access if relevant Create the Sampling Frame Determine the Sampling Procedure Design the Audit Sheet Define Data Collection Procedures Establish Timeline and Budget Set Up GIS Pilot Test the Audit Sheet and Data Collection Procedures Train Researchers Interviewers and Supervisors Conduct Auditing during Data Collection Analyze Data Hold DDM and Report on Findings STEP 1 DEFINING MAP STUDY OBJECTIVES When creating the study design for a MAP study either for coverage or access surveys the standard study design template should be used It is important to clearly describe the objectives of the stu
10. Access to condoms among female sex workers is 72 nearly three quarters of FSWs in the study area work or are based in venues that are within 200m of a condom outlet In the schematic representation below Fig 3 the minimum standard of access is defined as venues that are within 200m of a condom outlet The 72 from the result statement above would be those FSWs situated in the blue happy faces that represent venues with access So the FSWs in these venues are within the outlet catchment area Alternatively those venues represented by the red unhappy are considered too far away from the nearest condom outlets Thus these FSWs in these venues don t have access Figure Three Access to a Product or Service Delivery Point 6 outlet o When defining access remember the following The size of the catchment area is decided by the social marketers When possible the size should reflect existing marketing objectives in the same way that the product and marketing objectives help define the geographic unit of analysis and minimum standards of availability in coverage studies However marketers do not always set clear objectives in terms of proximity of outlets for specific target groups Therefore researchers should discuss the appropriate minimum standards for accessibility with relevant program staff m TRaC study findings or qualitative research findings on purchase behavior if available should be used to determine app
11. G DATA COLLECTION Auditing is the process of verifying whether an outlet or a service provider meets the required minimum standards of availability and quality The process is the same used in coverage 17 studies though access studies record the coordinates of each outlet and or hotspot with a GPS unit Also information on the population segment of interest is collected See the Coverage toolkit for detailed guidance on this step STEP 14 ANALYZING DATA The analysis of access data is mostly done in the GIS There are several steps in this process Step 1 Import attribute data on target audience sites including the number of people in each site Based on the latitude and longitude values each site will be displayed in the GIS along with its attributes on the type and size of the target population In the example in Figure 5 below the first meeting venue Rochets Road House is located at 27 315094 latitude and 31 888999 longitude It is a restaurant that is located in Emadladieni community in Shiselweni region Additional variables on condom availability target audience presence quality standards etc are included in subsequent columns not shown in this example Note that dBase dbf and Excel files xls or xlsx are the two main table file formats that can be read by GIS software If your original data is in SPSS then the relevant variables first need to be exported to either dBase or Excel Figure Five Example of a
12. Since people generally travel greater distances in rural areas to purchase products or to access a given service the standards will differ from those of urban areas HIV Counseling and Testing HCT sites for example may have a catchment area of a few kilometers in large urban areas mainly serving people from within the city HCT sites in more rural areas may cover a much larger region where the site attracts clients from further afield perhaps up to 30 kilometers away Quality standards such as outlet operating hours or absence of regular stock outs of the product of interest may be incorporated into the access definition For example access could be defined as the proportion of people in entertainment venues who are within 200 meters of a condom outlet that is open until at least 10pm and that does not report regular stock outs of condoms Analysis differs for general population groups As mentioned earlier the access indicator is particularly suitable for studies on the availability of products or services among specific high risk groups that congregate in well defined areas such as hotzones In a GIS it is usually more difficult to apply a catchment area to products that relate to general residential population groups e g youth women of reproductive age children under 5 etc because the analysis would require the precise location of every household in the study area This is the case especially when evaluating access to pr
13. al definition of what is meant by a hotzone before data collection begins This definition will vary from one country to another Then the field teams will define the precise hotzone boundaries on the ground at the start of data collection See Appendix 1 for several examples of how previous MAP surveys defined HIV high risk areas Surveys on access to services generally require larger areas such as districts communities or towns See the following section on minimum standards as well as the Coverage toolkit for more details on how to choose the type of geographic area and decide on an appropriate size Indicator Definitions and Minimum Standards Access Access is an indicator that measures the availability of a PSI product or service in relation to target population distribution It is defined as the proportion of a population segment in a geographically defined residential area or hotzone that is within the catchment area of a product or service delivery point As the catchment area is defined in terms of time or geographic distance this distance can be considered the minimum standard for access Note that the underlying assumption for using access as an availability indicator is that those people or communities who live work meet entertain within the catchment area of a product or service provider are more likely to use the product or service than those who are too far away A typical result of an access study would be phrased as follows
14. ap outputs Interviewers and supervisors also need to know how to use GPS devices ArcGIS trainings typically require five days after which participants are expected to be able to operate the software on their own display geographic information import attribute data analyze access generate maps and create GIS outputs maps charts data reports Ideally three to five staff receive this training primarily researchers but also MIS managers or monitoring and evaluation staff in programs that wish to apply GIS beyond simple access analysis for MAP studies R amp M also organizes regional GIS training workshops with a standard 5 day curriculum lt is important that trainees have adequate computer literacy before this training and that they will be able to practice their new skills with ArcGIS soon after the training Training field workers on how to use the GPS devices only takes about an hour They need to be taught how to retrieve the latitude and longitude coordinates of outlets and or hotspots A short GPS manual exists for the standard model that PSI uses in most platforms the Garmin eTrex device see image under Step 8 Most devices also come with a short user manual In addition to training data collectors on the technology they need to be familiar with the concepts of hotspot hotzone and target audience This knowledge will help them identify target audience sites and estimate target audience size if needed STEP 13 AUDITING DURIN
15. cts layer and mean is the access rate The above steps were used to calculate the proportion of households that are within 4km of an ACT outlet 8711 out of 5085 households or 72 9 Statistics of HH_all_UTM38 Field ACT_4km Statistics Count BOBS Minimum 0 000000 Masimum 1 000000 Sum 3711 000000 Mean 0 729794 Standard Deviation 0 444066 Frequency Distribution 0 0 01 0 2 0 3 0 4 05 06 OF 08 09 28
16. d a dbf table to ArcMap and use the function Add XY Data under the Tools menu Step 2 Use the Select by Location function to automatically select all records ina population layer that are within the catchment area of the product outlets To access this function go to the Selection menu and choose Select by Location Then specify the parameters as shown in the example below Click Apply after you finish selecting the parameters Select By Location f x Lets you select features from one or more layers based on where they are located In relation to the features in another layer Le The layer with the esc estes ton J POPULATION data the following layers HH_all UTM A C Qutlud Events MG_HH_s1 Events ka The layer with the PRODUCT M Only show selectable layers in this list data If the layer contains all that outlets and an attribute for availability first select only the features in this layer those records with value Yes AIL ACT outlets UTM 2 z or 1 Then check the option e aA Use selected features W Apply a buffer to the features in Al ACT outlets UTM38 of 4 Kilometers Pe ety The red features represent the features in All ACT outlets UTM 38 The highlighted cyan features are selected because they are within a distance of the red features Check the option Apply a buffer and specify distance radius and units km or meters Ws WSL Points L
17. dy and to understand how these affect sampling data requirements data collection procedures data management and analysis The primary objective of a MAP survey is always to monitor the availability of one or more products services within a given intervention area Most access studies will have the following specific objectives Monitor the geographic coverage of products or services Monitor the quality of coverage of products or services Monitor the penetration rate at the outlet or facility level Monitor the level of access to products or services among the target group of interest In studies involving more than one target population group equity of access may also be examined The first three objectives are explained in the preceding toolkit chapter access and equity of access are defined in the next step STEP 2 DEFINING THE UNIT OF ANALYSIS AND MINIMUM STANDARDS FOR ACCESS Unit of Analysis and Sampling Unit MAP for Access studies do not use geographic area as the unit of analysis which is used in Coverage studies Because the access indicator is always defined as the proportion of a given population segment that is close to a product or service provider the unit of analysis for access studies is the target population So in order to evaluate access and equity of access it is necessary to have data on the size of each population group Obtaining this data requires either a census of outlets and target audience sites wit
18. e of these realities it is necessary to simplify our access standards so they can be applied throughout the analysis This simplification also avoids the need for highly sophisticated spatial analysis techniques and much more detailed data than we currently collect Even so it can happen that it is not possible to record the precise location of outlets and target groups usually because no GPS units are available or because detailed population datasets are unavailable In this case a simplified definition of access can be used under which access is defined as the proportion of a population segment that lives in geographically defined areas or hotzones where the product or service is available n other words it is the proportion of people who live in areas where there is coverage see Fig 4 In the Coverage column of this table 1 indicates that the area was covered according to minimum standards while 0 stands for not covered 10 Figure Four Example of a Table Used to Calculate Simplified Access Rates Estimated Condom Population Province Population Coverage with Access O Nord Kivu Bishange Nord Kivu Buroha Nord Kivu Bweremana Nord Kivu Karuba Nord Kivu Kibututu Nord Kivu Kirotshe Nord Kivu Mapendo Nord Kivu Mitumbala Nord Kivu Rubare Nord Kivu Rubaya Nord Kivu Rugari Nord Kivu Rutshuru Nord Kivu Sake Nord Kivu Umoja Total 64543 Access Rate with Access Estimated Population 43910 64543 e
19. e target areas are precisely located The list of target areas should be as precise as possible and each hotzone should be labeled with a name and location information so that it can be easily identified STEP 6 SAMPLING Sampling in access studies follows the same procedure as in coverage studies see the Coverage toolkit for more information We recommend the use of LQAS when the number of geographic units is greater than 25 30 or so drawing a sample of 19 areas or zones in which an audit will take place When the number is lower then we usually conduct a census of all hotzones Access studies can also use purposive sampling with hotzones selected based on their programmatic importance In such cases no random sampling is done areas are selected through discussions with the programmers based on their relative importance or on how representative they are thought to be lf a census is used sampling is not needed Instead all product or service providers and all target audience sites in the study area are audited STEP 7 DESIGNING THE AUDIT SHEET The audit sheet for an access study is very similar to the questionnaire used for coverage studies However two important additional categories of variables need to be collected 1 Geographic coordinates of each outlet and each target population site These coordinates are defined by the location s latitude and longitude which allow for accurate mapping of each location in a GIS Data
20. ers may be 75 68 and 57 respectively In this case female sex workers have the highest access level But there is inequity of access 11 when FSWs access is compared with the access of other target groups MSM access is 7 percentage points less and truck drivers is 18 percentage points less When defining equity of access remember the following There is no unique cut off point to decide on whether there is equity of access or not However equity of access can be simply defined as a difference of less than 5 percentage points in access rates of individual target groups Equity of access can only be calculated if sufficient information about the various target groups is collected In addition to knowing whether a given target group is present it is essential to know the approximate size of this group Such information can be difficult and time consuming to obtain For these reasons many MAP for Access studies have not reported equity of access STEP 3 DEFINE APPROPRIATE STRATIFICATION MAP for Access studies also stratify data into Supervision areas as in coverage studies It should be noted however that many access studies do not include multiple Supervision areas Often access is measured in a sample of 19 hotzones nationwide as an additional component to the coverage survey which is generally conducted in multiple supervision areas The number 19 is derived from the LQAS method In such cases there are two distinct co
21. hin the sampled areas or a good population distribution dataset from secondary sources e g latest demographic census In the case of a census of most at risk populations MARPs researchers must estimate target population sizes for each site To do so easily it is recommended to select small rather than large areas as the geographic reference unit Because it is generally not possible to cover an entire program intervention area with this census researchers should at least sample geographic areas in which the study will be conducted These geographic areas are the primary sampling unit which needs to be well defined When defining the sampling unit for an access study remember the following The geographic area used as the sampling unit is usually the hotzone Because the majority of access studies investigate population groups at high risk of HIV infection such as female sex workers FSWs the hotzone is used as the sampling unit A hotzone is an area where there is a concentration of sites where high risk groups live work entertain or engage in high risk activity These sites are called hotspots These hotzones includes streets where sex workers are stationed a truck stop or a collection of bars in a neighborhood where members of the high risk groups meet They often correspond to priority intervention areas for projects involving MARPs and other highly localized target groups It is important to establish a gener
22. hout the month or season it is recommended that any questions on the target group size be formulated in a way so that the data can be compared for all sites An example of a good question on target group size is On an average Friday or Saturday evening or night how many clients come to this venue Note that this information on population size does not need to be collected in access studies among the general population when detailed georeferenced population datasets exist e g population by census enumeration area number of residents per village STEP 8 DEFINING DATA COLLECTION PROCEDURES The data collection process is the same in access studies as in coverage studies although access studies use GPS units to record the coordinates of each outlet and or hotspot These geographic coordinates must be recorded in the standard WGS84 format using decimal degrees as the unit of measurement noted as hddd ddddd in the GPS settings menu Specify these parameters when setting up the GPS units for the first time For example the coordinates of an outlet may read as follows on the GPS where N stands for North of the Equator latitude and E for East of the Greenwich Meridian Longitude N 46 18146 E 005 91526 In the field surveyors will collect data on the types and size of target population groups such as the approximate number of sex workers that are usually present in a hotspot during the evening Other
23. ines Polygons Apply Close 2 Step 3 Now create a new field in the table where you can copy the value of the population size of the outlet number of clients Only copy the value from those records that were selected under step 2 i e those that have access To do so open the attributes table right click on layer to see this option then click the Option button and choose the Add field Give an appropriate name to this new field e g Cond200m Then right click on the header of this new field to use the field calculator Calculate Values In the Field Calculator that pops up now choose the population variable in the Fields box by clicking on it In the example below field Q11C8 is the field with the number of clients in each hotspot Click OK after the field is chosen Field Calculator Fields Type f Humber C String 0 Date Advanced 11103 Step 4 Use the Statistics function to get access information about a specific site including the access rate To access this function right click on the header of the newly created field You will get the sum of the population that is within the catchment area of the products layer This number must then be divided by the TOTAL population in your population layer e g number of clients in all hotspots in the study sites In the example below the count is the total population the sum is the number of population members within the catchment area of the produ
24. information such as the estimated distance of villages to the nearest road or the opening times of outlets may also be collected As mentioned earlier access data collection on access indicators always involves a census of all outlets and of all target audience sites within the selected geographic areas The only exception is for studies on access to products or services that have a large catchment area In these studies it should be possible to use existing census and SES data provided they are detailed enough and can be integrated into the GIS In those cases there is no need to collect population size data 14 As with all MAP studies verbal consent should be sought from all respondents before beginning the data collection process Depending on local norms for research studies and the sensitivity of the target group or area in which the study is conducted researchers may need to obtain written consent using informed consent forms STEP 9 ESTABLISHING A TIMELINE AND BUDGET When planning for a first MAP for Access study programmers need to factor GIS set up into the timeline and budget See the next step for more information on the activities and technology involved in GIS set up The cost of installing a geographic information system which includes the software capacity building and a technical assistance visit by PSI s geographer is approximately 8 000 10 000 Alternatively platform staff with a GIS background and basic kno
25. l of these datasets varies tremendously among countries resulting in different levels of potential GIS applications For example if no digital maps of census enumeration areas exist it will not be possible to generate maps showing availability data at this specific geographic level or to use the variable population by enumeration area to estimate access to a given product or service Over the past few years R amp M has gradually built up a fairly comprehensive library of digital map datasets These have to be updated on a regular basis since administrative boundaries may change new information may become available e g population census or because existing basemaps may be improved e g road network As such it is always recommended to find out whether any updates are available from local GIS resources and other GIS users in the country or in the region These resources may include government agencies survey department national geographic institute ministry of health UN agencies or non governmental organizations Where limited public domain GIS resources exist it might be necessary to purchase digital maps from private companies though these are 1 Another reason for phasing out HealthMapper and replacing it with ArcGIS is that the former does not have well developed analysis functions and only contains basic layout features Moreover the software hasn t been updated in several years Note that HealthMapper is sufficient for acces
26. main difference between Coverage and Access studies resides in the analysis methods Access studies usually rely on spatial analysis techniques in a geographic information system GIS to identify areas with poor access among a given population segment In terms of data collection access studies go further than coverage studies by requiring that all sites hotspots outlets villages and clinics are georeferenced Georeferencing sites means that their geographic coordinates are retrieved in order to integrate data into a GIS for mapping and analysis purposes Coordinates are recorded with Global Positioning System GPS units which provide the latitude and longitude values of any site on the globe MAP for Access studies use GIS as the primary tool of analysis GIS can estimate the proportion of a given population group that are situated within a specified distance from a product or service delivery point They can also combine survey data with other geographic datasets Geographic Information Systems The use of GIS is key to the successful implementation of an access study On the simplest level GIS are computer based applications that are capable of capturing storing displaying analyzing and reporting spatial data However implementing GIS involves many things In this broader view GIS are a combination of hardware computer equipment printer GPS and other data collection tools software GIS applications database software procedure
27. mponents to the study 1 product or service availability in general areas of the country stratified by administrative division region province district etc and 2 access to products or services in specific high risk areas without any stratification levels Stratifying according to the type of area or to the level of risk may be more relevant than to Stratify simply by region Some studies specifically those with a focus on equity of access will measure access separately in two or more categories of hotzones For example researchers will measure access in hotzones where FSWs and their clients are present and in hotzones where a majority of venues are mainly attended by an MSM audience There is of course often a certain degree of overlap between target audience distributions so some hotzones may contain more than one population segment Another stratification option is to Stratify risk areas into different categories of risk e g low medium high risk or according to their size e g small medium large STEP 4 BENCHMARKING Benchmarking is used primarily for analyzing coverage and quality of coverage when LQAS is used It can also be used for analyzing equity of access Benchmarks are the values that define whether the coverage or quality of coverage results for a given supervision area are acceptable or not To a certain extent the same may be applied to the equity of access indicator To do so researchers must define the minim
28. n Attributes Table for a Dataset on High Risk Venues as Seen in ArcGIS Note not all fields are shown Table Ix cJ B l Bia ey nL _ Shape opsectip omame oaTyPE CATEGORY Lat Long REGION ozLocation community gt Point ZM 8 Rochels Road House Restaurant Food Outlet Formal outlet 31 888999 Shiselweni Urban Emadladieni E Point ZM 7 Moyamunye Butchery Butchery amp Brai Points Formal outlet 27 315671 31 891578 Shiselweni Urban Emadladleni Point zm 18 Nsalitie General Dealer Grocery shop Formal outlet 27 315016 31 643845 Shiselweni Rural Sitio Point ZM 20 Good Hope Bar Bar Formal outlet 27 312324 31 888466 Shiselweni Urban Emadladleni Point zm 28 Thabede Sheeben Informal outlet 27 288334 31 672978 Shiselweni Rural Sitio Point ZM 1 Lavurmisa Hotel Bar Formal outlet 27 316384 31 889778 Shiselweni Urban Emadladleni Point zm 2 Museni Spaza Spaza Shop Phone Outlet Informal outlet 27 316066 31 688874 Shiselweni Urban Emadladleni Point zm 3 Phetsile Investment Restaurant Food Outlet Formal outlet 27 315934 31 888999 Shiselweni Urban Emadladleni Point zm 42 A Nxumalo Sheeben Sheeben Informal outlet 27 270085 31 472829 Shiselweni Rural Hluthi Point zm 51 Shongwe s Sheeben Sheeben informal outlet 27 26264 31 535155 Shiselweni Rural Hluthi Point ZM 52 Kamahhishaza Sheeben Sheeben Informal o
29. n et al 2005 PSI and WHO have established a partnership agreement that outlines data sharing and capacity building rights and responsibilities and gives PSI access to the WHO HealthMapper HealthMapper is currently still in use in several PSI platforms However since 2010 R amp M has promoted more advanced GIS software ESRI ArcGIS Desktop which previously was only available under expensive licensing fees Through ESRI s non profit scheme PSI only pays about 100 per annual license for the full software package including advanced mapping and analysis extensions Identifying Digital Maps and Reference Datasets Any GIS relies on a series of digital maps often called GIS layers and on various reference datasets the attribute data such as demographic socio economic or public health data Digital maps include the boundaries of administrative divisions e g regions provinces districts census enumeration areas the location of settlements and of health infrastructure e g health centers clinics pharmacies the country s road network and environmental information such as rivers lakes altitude and land cover Other reference information includes population census data or detailed population estimates relevant public health indicators such as HIV prevalence rates by risk group or malaria incidence data and information on the socio economic status of populations It should be noted that the quality and level of detai
30. n in Figure 8 These maps can be valuable tools for designing and planning field interventions See also Fig 1 for another example Figure Eight Example of a Map Showing Access to Condoms and Lubricants Access to Lubricants vs Condoms Managua Nicaragua 2007 pa as LUGRICANT AVAMABLE IN OUTLET 2007 Yee 200MT LUBRICANT ACCESS D CONDOM AVAILABLE IN OUTLET 2007 ValucTexte Ye 200MT CONDOM ACCESS OUTLET 7 HOTSPOT o PRIMARY AND SECONDARY ROADS 21 Note that it may be useful to give programmers access to interactive maps rather than static images These can be created using Google Earth and or Google Maps two user friendly software programs that are ideal for creating and sharing maps with other users An example is shown in Fig 9 Data just needs to be in Excel format in order to use these tools See the PS Google Mapping Toolkit Duvall amp Piot 2011 for information on how to use Google Earth and Google Maps to create and share interactive maps Note however that these two tools are not GIS software and as such they do not have any analysis functions and cannot be used to calculate access Figure Nine Example of a Google Map Showing Product Availability Naine of Gullet naniwe grocery Type of Outlet Grocerdkiosk Condoms faible Chishanga Unbranded Source EBT Prev Project 57011 PSiiMalawi PSI Walawi website n Cep r CEY ae a oe _irn ET LE y pam gt
31. n the MAP methods and applications used in the different tyoes of MAP studies This information may be useful when presenting the tool to a colleague donor or other partner Overview of MAP for Access Studies MAP for Access studies integrate a geographic approach to product and service delivery system monitoring and take into account the concepts of need and target population distribution In this sense access studies can be seen as an extension of coverage studies Surveys on coverage and quality of coverage answer questions related to how widely a given product or service is available and how the product or service is made available to consumers Surveys on access go one step further by answering whether a given product or service is available close to the target population groups are we reaching the population segments among which need for the product or service is highest Even so MAP for Access surveys are very similar to MAP for Coverage surveys as they are usually based on the same sampling and auditing methods and they partially measure the same indicators Typically both MAP for Access and MAP for Coverage studies assess coverage quality of coverage and outlet penetration In addition to these indicators access studies estimate the level of access that people have to a product or service Such information is essential for understanding whether the social marketing program is appropriately reaching its target population groups The
32. nts of the report The sample tables below show how results from surveys on access to condoms among FSWs in four cities Table 1 and among three distinct high risk groups Table 2 would be presented in the summary report The examples are from studies that were conducted in Yunnan province China 2005 and in several provinces in Laos 2004 Table One Example of a Table with Access Findings Total FSW of FSWs within Access to HXH City population inthe the condom outlet condom among selected hotzones catchment area FSWs Kunming 6032 1385 Kaiyuan 1065 Gejiu Mengzi Table Two Example of a Findings Table on Equity of Access Target Audience Access Female sex workers 52 Mobile men e g migrant 47 workers or truck drivers MSM 32 The graph in Figure 7 is an example of how the data can be displayed Since it combines access rate by supervision area in the seven countries in the study and by target group MSM and FSW it is possible to compare availability levels between countries and between population segments 20 Figure Seven Example of a Graph Displaying Findings on Access Access to Condoms mFSW OMSM Researchers should also add maps displaying data to the report Maps can simply show the distribution of target audience venues or other locations where a specific population at risk is present along with product availability and catchment areas as show
33. o wish to implement a MAP for Access study It covers topics such as study design sampling data collection analysis and reporting Before reading this toolkit researchers should be familiar with the MAP for Coverage Studies toolkit referred to as the Coverage toolkit in this document The general background and introductory text are the same as in this preceding chapter so this section is not repeated here Briefly there are four types of MAP studies previously known as phases of Project MAP MAP for coverage and quality of coverage formerly Known as MAP phase This type of study aims to measure the geographic coverage of PSI products and or services as well as the quality of the distribution network measured by Lot Quality Assurance Sampling LQAS methods m MAP for access formerly MAP phase Il This study measures the level of accessibility and equity of access to PSI products and or services using geographic information systems GIS Combined TRaC amp MAP study formerly MAP phase Ill This study uses objectively defined measures of access to predict product and or service use correlating availability and behavior GIS applications for project planning and monitoring These MAP applications encompass the routine use of maps and geographic information systems to support decision making about PSI s social marketing interventions See also Appendix 1 of the Coverage toolkit for general information o
34. oducts or services with a relatively small catchment area less than 1km Analysis can be more challenging for products or services with a larger catchment area For certain products or services such as Long Lasting Insecticidal Nets LLINs or HCT it sometimes is possible to use detailed population data from the census by enumeration area or by village in order to estimate the proportion of the population that resides within the catchment area This detailed population data is often not available however In addition Knowing all the locations of the product or service delivery points in the study area can be another constraint Just as it is necessary to carry out a census of hotspots and of outlets when working in hotzones researchers would need to do a census of all product or service delivery points For example when estimating access to a social franchise network of family planning service providers one needs to conduct a census of all facilities under the network within the country or in a representative number of districts or regions at a minimum On the other hand platforms that have a well developed up to date and georeferenced MIS with service provider data and where detailed population distribution maps are available can calculate access with just these two pieces of information without conducting a specific field survey Access definitions are simplified All of the above catchment area examples are simplified representa
35. pots contain more than 95 FSWs and require at least 15 condom outlets Category B hotpots contain 30 95 FSWs and require at least 10 condom outlets Category C hotspots contain less than 30 FSWs and require at least 5 condom outlets Tanzania Hotspot a hotspot is any meeting place where people at high risk of HIV CSWs and their clients congregate or where high risk activity transactional cross generational or multi partner sex takes place A hotspot can be a bar nightclub or guest house Hotzone a hotzone is a closely clustered group of hotspots hotspots within easy walking distance from one another A hotzone is typically an urban neighborhood or street with many bars nightclubs groceries and or guesthouses Malawi Hotspot specific areas of high risk activity including truck bus stops bars taverns bottlestores lodges motels and nightclubs Hotzone the outer perimeter of a hotzone is ultimately defined by programmatic need Myanmar Hotspot specific areas of high risk activity which are different for each target population group Hotspots of FSWs include streets parks tea shops brothels massages beauty parlors hotels nightclubs and KTV restaurants Hotspots for MSM include beauty parlors bus stops nightclubs parks bridges street junctions teashops markets night markets and houses of Natgadaws Natgadaws are women believed to be chosen as consort by a nat spirit Sometimes these are men mainly MSM
36. pulation density by clan Pers_sgkm 1 7 8 10 11 15 m 16 21 iv BP ArcToolbox E Table Of Contents aja amp n g 8 855 7 749 Decimal Degrees OUTPUTS OF A MAP FOR ACCESS STUDY The primary output of a typical MAP for Access survey is a standard summary report that includes tables on product accessibility and other availability measures such as coverage quality of coverage and outlet penetration rates Additional tables with findings related to consumer price levels intention to sell reasons for not selling a product frequency of sales supervision visits and other variables may be produced as recommended in the Coverage toolkit Maps are also outputs of MAP for access studies that can be used as planning and decision making tools by program managers and field staff wno conduct sales and distribution service provision or communication activities Thus it is often useful to include a number of maps ina study report An example of a typical hotzone map with target audience sites and access to condoms is shown below Fig 2 Figure Two Example of a Map Showing Access to Condoms in a High Risk Area High Risk Venues and Condom Availability in Mbabane Swaziland 2009 Legend Condoms available No Yes High risk venues psi Catchment areas show access to condom outlets within 100 meters Data Source PSI 2009 MAP Survey pec 0 2 0 4 0 6 Kilometers HOW TO DESIGN AND IMPLEMENT AN ACCESS STUDY This
37. ropriate access standards A willingness to Typical Examples of Access Standards m Access to condom among MARPs 100 to 250 meters walk measure How far are you m Access to WaterGuard and other fast prepared to walk travel to purchase moving consumer goods used by the product X or to access service Y general population 500 meters to 1 provides valuable insights into realistic kilometer minimum standards for access Access to antimalarials hormonal contraceptives and other It is possible to use time distance as pharmaceutical products Between 1 an expression of access but it will need to be converted to its equivalent in geographic distance and 5 kilometers Access to HIV counseling and testing during analysis An example of time services family planning and other distance is a five minute walk to service providers 5 to 10 kilometers nearest condom outlet An example of an equivalent in geographic distance is Note that the above examples are largely the distance covered by a person based on expert opinion rather than traveling at average walking speed factual data However anecdotal evidence See the analysis section of this toolkit and certain studies e g Grossman 1986 for more information on this suggest that our common practices are in conversion line with true catchment areas and consumer behavior E Different access standards may be used for urban and rural areas
38. s data processing analysis and people surveyors GIS technicians decision makers By combining layers of spatial information and related attributes a GIS can detect and display particular distribution patterns or identify specific sites based on their location and on other user defined criteria Figure 1 below shows a screen shot of a GIS software displaying regional and county boundaries major rivers road systems and population densities GIS technology is applied to a wide range of domains and is becoming common practice within the areas of public health and human services Many PSI platforms have been increasing their use of GIS mapping for routine project planning monitoring and reporting Indeed the introduction of GIS as a routine management tool alongside traditional monitoring and evaluation M amp E tools strengthens Management Information Systems MIS increases data use improves targeting and planning and results in cost savings Figure One Combining and Viewing Geographic Information in GIS software ArcGIS Q Liberia_MAP_2011 mxd ArcMap ArcInfo File Edit wiew Bookmarks Insert Selection Geoprocessing Customize windows Help ce a Sk SR Kd ats te 1 2 050 426 llc E Gal Cel il ES bo e HAD 8 HHE op Se es GS Me Ea aii Table OF Contents ax a Leir Regions al 3 Major _Rivers 5 County aug M Road current CATEGORY Paved Primary Routes Track Tracks Po
39. s analysis 16 often very expensive Fortunately the need for purchasing maps is diminishing due to the increased availability of base maps on the Internet These GIS resources should ideally be obtained in an ArcGIS compatible file format usually shapefiles and any existing meta data about each layer should also be collected The meta data describe the contents accuracy source structure geographic reference system projection and any other relevant issues about the datasets STEP 11 PRETESTING THE AUDIT SHEET AND DATA COLLECTION PROCEDURES THROUGH A PILOT TEST Follow the standard pilot procedures that are explained in the Coverage toolkit In addition be sure to check the settings of all GPS units prior to sending out teams to the field and make sure that the devices have full batteries The units in which geographic coordinates are recorded should always be set to decimal degrees see Step 8 STEP 12 TRAINING RESEARCHERS INTERVIEWERS AND SUPERVISORS Again training for access studies resembles the training used for coverage studies The main difference is that researchers also need to know how to use the GIS software and how to operate GPS units Certainly they need to know how to collect spatial data They also need to know how to use the mapping software for integrating layers with product service and target audience site information By doing so they can analyze the layers and produce the access indicators and various m
40. sit several months later can be organized In addition it can take time to procure the technology needed for the GIS GPS units need to be ordered well in advance of when surveyors will be trained on data collection procedures Purchasing GPS units may take several weeks Before purchasing this equipment from the US or elsewhere programmers should inquire about any regulations on the use of these devices that may exist in the country If ArcGIS will be used as the GIS software note that the procurement process can take up to 6 weeks from initial order to final delivery of the software STEP 10 SETTING UP THE GIS To set up the GIS programmers need to procure GPS units and the appropriate software tool and research teams need to be trained on how to use this technology Building this capacity can take time especially in a platform that is new to GIS and mapping 15 As part of set up PSI platform teams need to decide which software to use as well as identify digital maps and reference datasets that are required to take full advantage of the mapping and analysis opportunities offered by GIS Choosing Software During the Project MAP pilot phase PSI Research amp Metrics R amp M worked to develop install and evaluate the cost effectiveness of HealthMapper This software is a simple customized GIS application for public health mapping developed by the World Health Organization for purposes of measuring access and equity of access Chapma
41. ta layer should show the outlet clinic locations with attribute data on availability and quality 18 Step 3 Verify the quality of the location data since sites and outlets with incorrect coordinates will negatively affect the survey findings Sites with incorrect coordinates will usually appear out of place and are easily identified The errors usually stem either from incorrect recording of GPS coordinates in the field in which case it may be difficult to correct the values or from incorrect data entry in which case one should refer to the paper questionnaires to retrieve the correct values Any sites for which coordinates cannot be corrected should be removed from the access analysis Once the required corrections have been made to the coordinates start again with Step 1 of the analysis process Step 4 Generate the catchment areas buffers around the outlets providers where the product service is available lf a time distance e g 5 minutes walk was chosen as the minimum access standard then this value needs to be converted to the equivalent in meters The standard rule is to use the conversion of 15 minutes 1 kilometer since the average walking speed is about 4 km hour Step 5 Identify and select the target audience sites that are within the catchment area of one or more product or service providers This identification can be done automatically through a spatial query select based on location in ArcGIS
42. tions of the true catchment area of an outlet or service provider In the real world the size and shape varies from one outlet to another depending on how popular an outlet is and on factors such as outlet densities and road infrastructure Moreover a catchment area s shape is rarely a circle road access physical barriers and administrative and cultural boundaries are all likely to influence an outlet s true catchment area It is important to keep in mind that we usually measure access based on a straight line distance rather than actual distance along roads or paths A given access definition can thus be seen as being little precise and corresponding to realities on the ground but because more precise access measurements are more complex to calculate and because we can apply the same minimum standard for access throughout multiple rounds and different studies we do not usually perform sophisticated network analysis to estimate precise access times and distances Indeed professional GIS software such as ArcGIS can calculate distances along road networks and travel time as well as incorporate physical barriers and other important factors when determining access to a health product or service provider Such an analysis is however more complex and relies on accurate and comprehensive road network datasets For more information on determining access to health services in general see Rosero Bixby 2004 and Noor et al 2006 Becaus
43. um standard for equity of access prior to the analysis stage In other words they must determine what level of difference between the access values for each population group is acceptable for there to be equity of access e g 5 See also Step 2 above on setting minimum standards This benchmarking step is only applicable when it is possible to calculate access levels for more than one individual target group it has not been done systematically in the past 12 STEP 5 CREATING THE SAMPLING FRAME As in coverage studies access studies require a complete and accurate list of geographic units in order to select the study sites Readers should refer to the Coverage toolkit for general information on the sampling frame For access studies in particular the geographic sampling units will be the so called hotzones in most cases aS opposed to census enumeration areas or villages Comprehensive lists of such high risk areas are generally not readily available As a result researchers need to create a sampling frame which is a list of all known hotzones from which a random sample will be drawn It is often possible to combine information from the Behavior Change Communication BCC and sales departments from previous MAP surveys if any were conducted and from external sources such as other NGOs that work in the same target population sites PSI field staff such as sales representatives and peer educators often have very good knowledge of where thes
44. utlet 27 260087 31 437059 Shiselweni Rural i Mgazini Point zm 53 Mahhishaza 2 Sheeben Informal outlet 27 259873 31 437301 Shiselweni Rural Mgazini Point ZM 66 Diamini s Sheeben Sheeben Informal outlet 27 254564 31 587713 Shiselweni Rural Nsingizini Point zm 98 Mngometulus Sheeben Sheeben Informal outlet 27 217303 31 65704 Shiselweni Rural Hlushwana Point ZM 118 Sebenzile Restaurant Restaurant Food Outlet Formal outlet 27 20336 31 556374 Shiselyeni Rural INsingizini Point zm 142 Maphango Sheeben Sheeben Informal outlet 27 497778 31 333366 Shiselweni Rural iNgwenyameni Point zm 181 Phumlani Grocery General Dealer Grocery shop Formal outlet 27 164974 31 779648 Shiselweni Rural Mdlanyoni Point zm 199 41 Bar Bar Formal outlet 2714606 31 667954 Shiselweni Rural Nkutjini Point zm 205 Enhlanhleni Wine amp Malt Bar Formal outlet 27 161153 31 266629 Shiselweni Rural Thembelihle Point zm 217 Msibi Sheeben Informal outlet 27145376 31 512118 Shiselweni Rural Ka Mngayi Point zm 230 Kunene Sheeben Informal outlet 27144547 31 124062 Shiselweni Rural Ngelane Point ZM 235 KaMsibi Sheeben Sheeben Informal outlet 27 142815 31 124519 Shiselweni Rural Ngelane m 5 gt 4 1 gt pt B 0 out of 612 Selected Step 2 Import attribute data on product service availability in a separate layer This da
45. wledge of access studies or experienced local consultants may be able to assist with this process Equipment costs are another significant portion of the budget GPS units cost approximately 110 each and each team of interviewers needs one unit in the field Despite high installation costs PDAs are a cost effective alternative for paper questionnaires and traditional GPS units Several PDA models and many smartphones have an integrated GPS receiver or can handle an external GPS extension Please refer to the PSI Research Toolkit chapter on PDAs Piot et al 2006 for more information on the set up process and cosis for implementing GPS on PDAs Besides these GPS units a computer and a printer preferably color no other equipment is required to set up a GIS The timeline to implement an access study is similar to that of a coverage study For platforms that are conducting an access study for the first time it is recommended that the PSI geographer does a technical assistance visit during the study design phase He will help set up the GIS collect and organize basic geographic information and build capacity among researchers This visit may be combined with a broader GIS scope of work so the platform can broaden its GIS capacity further Training on data collection methods in particular on the use of GPS units is also included Depending on the program s need for further capacity building or for assistance in analyzing data a short follow up vi
Download Pdf Manuals
Related Search
Toolkit_MAP_for_Access_Studies