IMapp-Modelling Immunisation Outreach in Far-flung Rural Clusters

By: Dr Manosi Lahiri
IMapp, the Immunization App, was developed to strengthen the delivery of public health services, taking advantage of an innovative multidisciplinary approach and digital technologies like Geographic Information Systems (GIS), Remote Sensing (RS) and Machine Learning (ML). The initiative demonstrated the suitability of porting public health data on a GIS platform to get good insights into the distribution of vaccines; create an automated indexing system to monitor the efficacy of the Universal Immunisation Programme (UIP) on the ground, and geographically extend vaccination to include all rural population clusters. It was validated in the East Champaran district of Bihar, India, in the years 2019 and 2020, under the initiative Immunization Data: Innovating for Action, supported by Grand Challenges India.
Health Population

Introduction

A Delhi based company, ML Infomap has currently developed an Immunisation App, IMapp, to strengthen the delivery of public health services in an innovative manner, taking advantage of digital technology. The Universal Immunisation Programme (UIP) of the Government of India (GoI) routinely collects and stores data on vaccination given to infants, children and pregnant women (Universal Immunisation Programme, 2012-2013). Although 98 per cent of districts report monthly data to the nations’ centralised Health Management Information System (HMIS) since 2009-2010 (Rural-Health-Statistics, 2019-2020), even a cursory glance through public health data indicates that the distribution of vaccines and the reach of the immunisation programme is uneven in rural India, and pockets are left out of the vaccination net (Fig. 1). It is important to understand the reasons for this, especially since they vary from place to place. The core aim of our initiative was to geographically extend vaccination to rural population clusters in peripheral areas and create an automated indexing system to monitor the efficacy of the UIP on the ground.

Fig 1. Number of Women Registered for Antenatal Care in Bankatwa block of East Champaran district. Hamlets and villages indicated by green symbols do not have any women registered. Many of these places are located far away from paved roads.

Collection of health data 

Health data is collected by state governments, aggregated, and maintained in several different systems by the Ministry of Health & Family Welfare (MoHFW). Besides, these systems are specific to some states and do not cover all parts of the country. For instance, the Electronic Vaccine Intelligence Network (eVIN) covers 12 states (Vikaspedia, undated). Very often, the data and analytics required by public health officials are beyond their reach for many different reasons. For example, the data needs to be accessed from several different institutions, data is not interoperable, the official is not familiar with the techniques to analyse the data and perhaps has time constraints as he has just too many other functions to perform. Yet, he does need to understand what the data indicates and take actions based on his conclusions from them.

The process of data collection is outside the purview of IMapp’s engagement and its pilot. IMapp is an automated method of interfacing, modelling, mapping and reporting the immunization coverage and vaccine consumption data. The purpose is to ease the work of the health officials in the block and district and to enable all state health officials to get a comprehensive overview of the immunisation programme being conducted across the state. The wide adoption of IMapp will be in the interest of distributing vaccines universally and equitably.

In the pilot conducted in Bihar, data from HMIS and eVIN were complemented with topographical maps from Survey of India, infrastructure data from government records, satellite images of different spectral, temporal and spatial resolutions, Census demographics, and more, to understand the complex nature of vaccine distribution in East Champaran. Our experience in geospatial technology enabled us to extract meaningful location intelligence and demonstrate the importance of gaining insights into the problem of equitable vaccine distribution.

 Analysing health data for equitable immunisation

Focussing on `Immunization Data: Innovating for Action’, IMapp is an automated system for routine analysis, visualisation and reporting of health data to spread immunisation equitably across all geographies by:

  1. triangulating the immunisation coverage data with the vaccine consumption data to understand the nature of vaccine distribution from the cold storage point to the vaccination session sites;
  2. tracking immunisation coverage indicators of mother and child population included in vaccination sessions, and;
  3. providing a geospatial platform that enables ease of access to actionable health and related data and analytics.

East Champaran district in Bihar was piloted to test the efficacy of IMapp and counter-validate test results generated from the app (Fig. 2). This is a populous district with low development indicators. The Bihar State Health Society (BSHS) was interested in assessing the utility of digital technologies in impacting the immunisation programme here and in similar districts.

Fig. 2. Map showing performance of health centres in conducting vaccination. Areas represented in green have done better than the areas marked in red.

Digital connectivity 

As IMapp is designed for office use and is not a mobile app, it is aimed to enable public health officials to make informed planning decisions using the health data generated from their areas of responsibility. As both digital connectivity and digital literacy are challenges in the rural parts of the country, a hybrid online-offline system has been developed to counter the intermittent electric power blackouts and patchy telecom network. IMapp enables the block health official to download the data files on his desktop and undertake his routine work offline, and later email the report to the district or state office.

As IMapp’s web interface is simple and intuitive to navigate, it is easy for the officials to operate it. IMapp actually helps in reducing workload and is designed to help make selections from dropdown lists. The block health official only needs to know what indicators to use and for which districts or blocks. Calculations for generating indexes and mapping are fully automated. Advocacy, training and support are part of IMapp’s deliverables.

In addition, field workers having access to digital devices allow administrators to utilise their services in improving the quality of data by collecting digitally from the vaccination session sites. With rising awareness of the possibilities among senior officials, there is an enhanced scope to provide robust software support to field workers in taking immunization to remote places that were ignored earlier.

 Results from the East Champaran study

The present findings are from the initiative Immunization Data: Innovating for Action, supported by Grand Challenges India, funded by the Bill and Melinda Gates Foundation in partnership with Biotechnology Industry Research Assistance Council (BIRAC), GOI.

The East Champaran district is approximately 4000 sq km. This district was selected by MoHFW and BSHS for conducting the pilot. In Bihar, immunisation coverage data is recorded from the vaccination session sites and collated for administrative blocks, districts and state and stored in HMIS. The vaccine consumption data is collected for the last cold chain storage point at the Community Health Centre (CHC) or Primary Health Centre (PHC) in the block or district and aggregated for the district and state in eVIN (Vikaspedia, undated). Both HMIS and eVIN data are updated monthly. The two systems are in independent silos, and the data are not directly interoperable. That is a challenge for the public health officials who need to validate the vaccination coverage data and monitor the immunisation programme across the state.

To undertake the distribution of vaccines across territories requires an understanding of the complexities of the area where it is to be delivered and consumed. The purpose of using a GIS platform was to meet our objective of identifying areas that are underserved in the immunisation programme and to develop a method of finding similar areas elsewhere. A workflow was designed using GIS tools and methods, including Machine Learning. This reduced the workload relating to several processes like analysing multiple satellite images, synchronising different resolutions and scales, performing analysis across several spatial layers, and avoided errors arising in the process.

Using satellite data, models were developed to identify low-lying areas prone to seasonal waterlogging and to extract the non-vegetation (buildings) in the image to emphasise clusters of hamlets not represented in the HMIS, eVIN or Census data, and which consequently remain outside the UIP net.

 Modelling immunisation/vaccination coverage 

IMapp was still in its ideation form when ML Infomap became a grantee of Grand Challenges India 2018 and subsequently the pilot was conducted in 2019 and early 2020. Its significance in planning for pandemics was not factored in. IMapp’s aim was to develop the app for data relating to immunisation of mother, infant and pregnant women and so focussed on vaccines like DPT (diphtheria, pertussis, tetanus), Japanese Encephalitis, Polio, BCG, Rotavirus vaccine, etc.

Vaccine distribution involves a cold chain and must be undertaken within a stipulated period to maintain potency and taken to the target consumers in a timely manner. Consequently, you need to have a very clear notion of where the consumers are, their numbers, accessibility and health infrastructure available to be prepared for any eventuality. As this collation of information takes time and skill, it needs to be a continuous exercise and available in times of epidemics or pandemics. In those places in the districts where routine immunisation programmes are already well in place, this becomes easier. Even for Covid vaccine distribution, the logistics are no different, though the volumes are enormous, both in terms of supply and consumption.

It is therefore important to triangulate the coverage and consumption data to evaluate the efficacy of the implementation programme and to improve vaccination reach to all rural population clusters. Many of these settlements are not listed in the published Census records and often ignored in vaccination drives because they may be prone to flooding or waterlogged or inaccessible by road. As IMapp addresses these issues, it can be used in normal times and in pandemic situations.

Conclusion

IMapp can be scaled up to include extensive territories, in both rural and urban areas. The execution of the pilot demonstrates the utility of using a GIS platform to solve the problem of accessing, modelling, visualising and analysing immunisation statistics from multiple systems. Maintaining continuous interactions with local health officials is essential to spread the benefits of using digital technology. Advocacy should be undertaken in the districts and blocks and not limited to state headquarters.

References

Rural Health Statistics. 2019-2020.  Ministry of Health and Family Welfare, Statistics Division, Government of India. Available at https://hmis.nhp.gov.in/downloadfile?filepath=publications/Rural-Health-Statistics/RHS%202019-20.pdf 

Universal Immunization Programme. 2012-2013. Ministry of Health and Family Welfare, Government of India. Available at https://main.mohfw.gov.in/sites/default/files/5628564789562315.pdf

Vikaspedia. undated. Available at eVIN Project of Health Ministry — Vikaspedia

Leave a Reply

Your email address will not be published. Required fields are marked *