Health Access Index (HAI)

• Primary Health consultations are often avoided due to multiple reasons : Lack of financial access, health center being too far, too busy etc. Even in the presence of public health facilities providing free of cost Healthcare (sometimes highly subsidized), they are still underutilized due to opportunity cost of time. People usually do not trade off a working day for a primary care visit, and only seek care when the problem has escalated to secondary or tertiary levels of clinical severity. 

• This is especially true for middle and lower socio-economic groups where the value of earning immediate daily wage is higher than long term benefits of timely primary care. 

• Our Methodology identifies cities that have ensured that primary care is accessible within 15 min from any location in the city, so that no one needs to make the choice between losing daily wages and accessing timely primary care.

• We work at the intersection of location intelligence and public health. We are currently working to enable sustainable primary health access in cities. We look at satellite images of cities and quantify primary health access as reflected in city design. 

• The logic is that if two cities are being compared for primary health access, the one designed better for the accessibility metric has a higher probability of better utilization of primary care services, ceteris paribus, thus leading to better primary care. 

• The underlying hypothesis is that people will utilize primary healthcare services, if any built-up location in a city ( homes, offices, etc.) is within 15 min of a primary health access point. It also needs to be connected adequately by public transport. We use satellite imagery to identify zones in a city which lack access to a public primary health access point within 15 min travel time. We call the metric - Health Access Index (HAI).

• HAI = fraction of city built up area within 15 min of travel time to a primary health access point. Closer the value is to 1, the better access the city can deliver, thus leading to higher utilization of primary care. 

• Areas of the city outside the coverage for access are considered underserved zones. These underserved zones when identified, can be then used by local municipal governments to take three types of decisions : 1) Plan to build a public clinic in the underserved areas 2) Plan public transportation routes that support connectivity to the nearest clinic ensuring 15 min or less, travel time 3) Incentivize private sector to open clinics in such underserved zones. As our cities dynamically change in population size, infrastructure composition and morphology, regular monitoring of zones is essential to the sustainability of primary health access in the cities. 

• Currently we are piloting the metric with two cities : Ambala and Panchkula in the state of Haryana. Once established, it will be developed into a web-based Geospatial application which can be used by any municipal government to map their health infrastructure and identify zones within the city based on health access.

• Our Data needs for the model are frugal and almost always publicly available. Data collection is being done in the following ways. 1) The Satellite imagery data of cities is free to use Earth Observation (EO) imagery from European Space Agency’s Sentinel-2 Platform. 2) The location of Public Health centers is available on the website for each State Department of Health. 3) The Public transportation routes data is being sourced from State Transport Department. 4) Traffic Data is being sourced from Google Maps Analytics.

• Additionally,  we are developing this geospatial application using  open-source software, to enable wider acceptability & adaptability. The Data collection does not require on ground survey or any high resource data collection. All the data is available on Public Sources. Additional data points like mapping the private sector will be done in later stages based on business registration data for private healthcare providers who open clinics with licenses in the city. Any Municipal govt can easily provide these addresses as data points. 

• The novelty of this method is the following. 1) It is extremely Data light and uses Public Sources of Data 2) It measures Health delivery input parameters ( Access to primary health infra) as opposed to traditional output or outcome measurement techniques. 3) It provides insights for action at the city planning level which impacts multiple lives. 4) It brings to fore, the location intelligence paradigm of primary healthcare access which is expandable to include data streams such as public transportation nodes, socio-economic stratification, demographics etc.

• 

The low and middle income social class are the first to avoid primary care for the opportunity cost of daily wages of a working day. Increasing access to healthcare within 15 mins of their location, will ensure that they do not miss a working day and yet have access to primary care when they need it. Currently they avoid primary care and reach the health system only when the problems are aggravated in clinical severity, leading to not only a higher cost on the health system, but also greater Out Of Pocket expenditure(OOPE). People at the lower end of the social ladder are more inclined to decide in favor of immediate gains in wages over long term benefits of health.

• Aditya was raised in Panchkula (one of the 2 initial pilot cities) and knows the local contours of the place. He’s an experienced hand at geospatial data applications, GIS-aided decision support systems, having honed his skills while working on diverse projects professionally and also during his Master’s degree at Georgia Institute of Technology, on a Fulbright Fellowship.
• Dr Debanshu Roy is a Public Health expert with a wide variety of professional and educational experience, which includes a Master’s in Public Health from University of Texas, on a Fulbright Fellowship.
• Sparrow Analytics has been a trusted partner in location intelligence and public health use cases for the State and local govt through multiple engagements in the past. We have worked closely with the local health administration during management of COVID through our solutions in GIS based COVID management. This has earned us the trust of the local and state level government leadership.
• Additionally, our experience of working on ground with public health officials rigorously during COVID , has enabled us to look closely at how the government departments function.
• This on ground understanding of the middle class in Panchkula, by being a stakeholder in the same group, as well as trust earned within the local govt for location intelligence and Public health, gives us a unique position to gather the required data, and develop and deploy the POC.

Our expectations from the challenge are two fold :

1) Financial : Most govt depts do not fund experimental software development and want pro bono use for a year before they can contract it for the next year. Once we detail out the methodology, we need to offset the cost of development and maintenance of the software solution for a year to prove utility of the same. 

2) Partnerships : The same methodology can be used to create Health Access Index (HAI) for all cities in the world as a Global Annual Report. This can be achieved through partnering with different organizations responsible for data in different countries. This will provide a competitive yardstick for cities to compete at the health planning level to get high ranks on Health Access Index. BMGF can be a critical partner towards the same. 

The Global City Health Access Index is envisioned to be a knowledge nudge tool with similar impact on catalysing the policy ecosystem for health, as the ease of doing business reports do for business. Cities which score higher on the index, will attract better citizens and companies and will thus benefit economically and culturally from their policy efforts. 

• The novelty of this method is the following. 

  • It assesses Primary Healthcare Outcomes as a function of Primary health Planning (as opposed to traditional output or outcome measurement techniques), a unique dimension often overlooked in health delivery. 

  • It has a very high potential for impact as it moved governance and policy initiatives, thus the scale of population being impacted is massive

  • It is extremely Data light and uses Public Sources of Data . 

  • The Methodology is easy to understand and cannot be disputed scientifically. 

  • It uses GIS technology and location intelligence to stimulate health delivery on ground. This type of approach has never been used by local government departments actively. 

  • It is a dynamic metric and changes based on govt interventions done in the area, thus can be used as a monitoring tool for policy interventions

  • Additionally, as residential clusters change based on increase in population, influx on migrant labour and other factors aiding urbanization, the zoning will dynamically change as well.

  • The chances of success are higher, since the metric is pivoted on Human centered Design. By assuming a travel time of 15 min to access healthcare, we keep human tendency to avoid primary care for daily wages at the center of our argument towards utilization of primary care services. 

  • The solution is highly scalable in Adoption and Scope. a) Adoption: Once the solution is built as a web-based interface, any municipal govt in the world can identify their city, upload their infrastructure and transportation data and get immediate zoning results. b) Scope: Once the solution is built as an online tool, any other datasets like DHIS2, PHCPI etc can be overlayed on top of it, to see the spatial patterns in these datasets.

• We intend to at least evaluate 2-3 cities next year and then expand to the 25 smart cities identified in India under the Smart Cities Mission of the Ministry of Housing and Urban Affairs in India. Doing so will set regional benchmarks for other cities in the vicinity which can then develop aspirational goals towards their urban planning activities. 

• Simultaneously we intend to evaluate 5 major global cities in the next year with local partners providing data on health centers to start building the foundations for the Global Urban Health Access Report.

• There is also a maturity roadmap for the HAI methodology based on subsequent data layering. The first version will use limited data sets. Further versions will include addition of region wise traffic data and public transportation routes, socio-demographic and population data, private sector location data etc as per availability. These data layers with partners will increase the richness of the index, highlighting other areas of interventions which the cities can adopt to increase Primary Health Access in their respective locations. 

• Outcome Level Indicators for success:  For participating cities, Higher city level HAI scores YOY,  YOY decrease in underserved zones within cities, and YOY increase in public transportation routes and frequency to health posts within cities.

• Output Level indicators of success : No. of cities added to index YOY, No. of new datasets added to HAI index methodology, No. of partners involved in implementing the HAI in their local city governments.

• The basic hypothesis is that people will utilize primary healthcare services, if any built up area in a city is within 15 min of a primary health access point, and connected adequately by public transport. We use satellite imagery to identify zones in a city which lack access to a public primary health setup within 15 min travel time.

• We are measuring the inputs to health services by measuring primary health access. If cities are designed in a manner where access to primary health is not time consuming, is affordable to reach via public transport and does not make the individual choose between daily wages and care, then we are creating enabling ecosystems for access to primary care. This is critical to ensure sustainability of cities with population increasing at a faster pace than health services supply. 

• Although there are multiple inputs and activities in the log frame model, we can focus on the main ones here for discussion.

• Activity : Measure Primary Health Access Index. 

• Output : Identify and highlight the Underserved zones within the city

• Outcome : City administration can make three types of decisions : 1) Plan to build a public clinic in the underserved areas 2) Plan public transportation  routes that support connectivity to the nearest clinic ensuring 15 min travel time 3) Incentivize private sector to open clinics in such underserved zones. As our cities dynamically change in population size, infrastructure composition and morphology, regular monitoring of zones is essential to the sustainability of primary health access in the cities.

• Earth Observation (EO) imagery from Remote Sensing Platforms are accessed, to get a wide area perspective of land cover (Habitations, Farmlands, Waterbodies etc)
• Geographic Information System (GIS) is used to bring all the diverse data streams onto a unified spatial interface.
• Machine Learning (ML) algorithms are used to identify land cover and its extents.
• Cloud Architecture is used to make this web-application accessible on any number of devices, to any number of users.
• Big Data architecture enables us to handle multiple streams of datasets, enabling it to adapt to even bigger datasets.

• All Data needed for the base version is public Data Sourced from open sources or Public Departments of health and urban planning. 

• The addresses for all public health facilities are already posted on the Department of Health website. Sometimes we re calibrate the location from ground partners (government and non-government). 

• We have multiple-domain experts working on the team that gives us a holistic blend of points-counterpoints to any problem statement.
• Since the time of our team’s inception as Sparrow Analytics, we have been focusing on developing solutions to issues that can have a societal impact. Public health, Improving governance and service delivery, access to clean water etc. are some of the areas we have worked in.
• Our company has a ‘Universal Equality’ clause in its charter that prohibits discrimination based on gender, ethnicity, race, religion, language, sexual orientation and so forth.
• Our company’s ‘Honor Code’ seeks to foster a spirit of amity & kinship within the team and in communities or business environments, where we operate.

Our Business model is to work with local municipal governments and non profits working towards urban sustainability and health, and providing them the model, and tool to plan for their cities. The revenue will be generated on a SaaS model where they can use the tool to map their services and add additional data layers as needed.

The Global Health Access Report is envisioned to be a separate effort which we can raise grant funding for the first couple of years and then nest it within a relevant multilateral Agency which will be the owner, while the tool can be used by different cities to develop their scores YOY. 

We are still evolving separate business models and will experiment with a few, and are bound to work with multiple channels on the same. 

It is envisioned as a mix of grant funding for the Global Health Access Report, as well as revenue from SaaS for the software planning tool being developed. 

The SaaS product is aimed not only on the local municipal governments, but can also be used by non profits in the public health space to map out access to their services or projects. This is envisioned to be a sustained revenue stream.

We have already received grants from MIT Solve for our other GIS based solutions in Public Health and COVID management. The response from the state govt stakeholders have been very encouraging and has led to several other GIS based projects.

We have grants from the State Government in deploying GIS based tools in multiple use cases eg: Department of Sports, Department of Environment. These grants are being used to develop the POCs for the use cases defined, post which they will also be developed into SaaS products.

Our Revenue streams are varied :

1) Grants for developing POCs

2) GIS based consulting projects with Government and Non Government organizations

3) GIS software development

Based on the stakeholder and partner, financial models are established.