This is a case study about how strategic use of data helped curb a huge public health challenge – the spread of HIV in India. Preventing the spread of HIV is recognized globally as one of India’s big public health success stories. Yet, within the public health community in India there is little sense of ownership of this success. At the centre of this story is the strategic use of data. We have documented this to help us imagine how we can replicate this success in other public health initiatives such as fighting COVID.
Introduction
While HIV in India is well hidden, the country has the third highest number of HIV-infected people in the world, at 2.1 million.[1] At the height of the epidemic in the early years of the millennium, there were about 2.5 lakh new infections every year. Even though the epidemic was concentrated among high risk groups such as sex workers and the transgender population - the sheer size of the country meant that HIV in India was cause for global concern. Prevalence among sex workers was over 5%, and there was a fear that this could easily spread to become a generalized epidemic. This did not happen. The government worked in hand with donors and NGOs to turn the situation around so that by 2012 India achieved a 57% reduction in prevalence.[2] By 2016, there were only 80,000 annual new infections. This was the highest rate of reduction achieved in the world. At the same time, India has also been able to provide access to treatment for 75% of those infected.
These successes occurred in a context where health was not a political priority, was chronically underfunded, the concerned communities were marginalized, social values were conservative, community participation was limited and the health system had weak organizational capabilities.[3]
What then caused the turnaround? Strategic use of data was key. Health is a knowledge intensive sector, that requires active generation of evidence and information. The National AIDS Control Organization (NACO) and its partners adopted many strategic data-driven approaches to make limited resources go further (see Table 1). India’s HIV innovative strategic information initiatives made the country a global front-runner in HIV prevention, and the success deserves notice for how these strategies can be replicated across other priority health areas – such as the COVID response.
Here is the story of how this happened. The success of the approach can be mapped through three different strategies: 1. Innovative Applications 2. Innovative Data Sources and 3. Building Capabilities. These are depicted in Figure 1. This case study has been compiled by a review of the available literature and interviews with key stakeholders involved with the process, who also reviewed earlier drafts to clarify and refine the description of the processes.
Figure 1: Data use approaches
1. Innovative Data Sources
The high level of stigma and the clandestine existence of groups such as transgenders and sex workers meant that risks were not identifiable from traditional data sources, such as household surveys. This meant some creativity was required to identify and target at risk groups. Innovative data sources and research methods were established and used by NACO to identify and track risk behaviors and populations. Some of these data sources are described below.
1.1 Community Mapping
Community mapping using qualitative snowballing techniques were deployed to identify high risk behaviors such as transgender sex workers across different geographies. Such mapping was done by professional organizations (such as Karnataka Health Promotion Trust and IHAT[1]) with the participation of members of the at-risk communities,[2] and validated by IIPS Mumbai. This information guided micro-planning, a process of targeting approaches to the specifics of the local context. The maps were aggregated to allow higher-level macro planning. Maps need to be regularly updated to capture the changing nature of dynamic communities.[3]
1.2 Individual tracking data
Tracking cards were used by peer educators to track contacts with each individual, and plans were developed to ensure all populations were reached. This data helped identify “opportunity gaps” to see which FSWs (for example) were lacking outreach services. Maintaining this system helped build the confidence and skills of peer educators – eventually helping them to become leaders in their communities. Data was entered into an MIS, and the aggregated level was also used to track coverage to plan outreach in each district and state.[4] This data was also linked with clinic data. Linking outreach data with clinical reporting data allowed partners to understand their performance and use their data for decision making.[5] Using mapping, tracking and clinic data for micro-planning allowed implementors to refine outreach and service delivery practices.[6] This innovation came from the field and was eventually adopted at the central level by NACO.
1.3 Quality monitoring
The common minimum standards provided a basis to monitor quality through different facilities and service delivery sites. In addition, exit interviews were conducted once every three months with a minimum number of patients, to allow continuous quality monitoring.[7]
2. Expanding Capabilities for Strategic Use of Data
While NACO was evidence-driven from the outset, the onset of “NACO 3” in 2007 (the third project phase) involved the scale up of all programs from high prevalence states to the whole country. There was a need to generate the required data for targeted programming at the district and sub-district level, in every district in the country. This required a massive expansion of data systems, and more data suddenly became available for analysis.
How did NACO manage to make use of this data central to its operations at scale? There were a number of key factors that allowed NACO to adopt innovative data-centric approaches to programming; including strong human resources, donor support and a vertical administrative structure that allowed teams to work independently from established government processes. However, capacities at the state and district levels also had to be built. This was done through a number of different mechanisms, described below.
2.1 Workshops
Poor capacity at the state and district level required capacity building workshops to use data to prepare plans, assess the needs of vulnerable populations and how to interlink activities. Public health institutions such as universities and research institutes were identified in each state to provide ongoing research and data use support.
2.2 Technical Support Units
External partners were key to driving data use by the State AIDS Control Societies. Initially state governments lacked capacity to scale up AIDS prevention programs or absorb donor funded initiatives. NACO contracted NGOs and technical assistance agencies to run state-level Technical Support Units to provide collaborative support in data use and program management (“in-sourcing” support). There was also a national level TSU run by Indian Health Action Trust (IHAT). This model helped improve data systems for performance tracking ensuring that national reporting formats were adopted. For example, KHPT’s TSU would provide Karnataka State AIDS Prevention Society (KSAPS) with analyzed data every month. This allowed the state to see the gaps in coverage, and adjust their strategies accordingly. [1] The TSU also provided a conduit to transfer best practices from KHPT’s own implementation into the KSAPS-led districts. The TSU model was also deployed by the Bill and Melinda Gates Foundation to address immunization and by the WHO to fight polio.
2.3 Supportive Supervision
Surveillance data is prevalence estimates based on blood tests from both pregnant women during routine ANC care and testing of high risk population groups. Prior to 2006, timelines, inspections and validations were hardly reviewed, and there were only 22 supervisory teams in the whole country. Additional testing sites were added in 2007 through a network of institutions including; All India Institute Medical Science (AIIMS), National AIDS Research Institute (NARI), Post Graduate Institute, Chandigarh (PGI), the National Institute of Epidemiology (NIE) and All India Institute of Hygiene and Public Health (AIIHPH), along with teams of data entry operators, epidemiologists and research analysts. Initial review of sites was alarming. Vacancies, lack of supplies and poor supervision lead to a tendency to “fudge” the data. Site-wise deficiencies were documented and addressed through training programs, guidelines, protocols, checklists, and reviews. Once the quality of surveillance data improved, prevalence estimates actually reduced from 5.7 million to 2.5 million.[2]
2.4 Annual Partners’ Meetings
Annual partners’ meetings at the state and national levels allowed for data to be intensely analyzed and discussed, facilitating a consensus on what the data meant, and what the implied action items should be.
2.5 M&E Infrastructure
One of NACO’s partners, Karnataka Health Promotion Trust, built in a strong M&E system that provided regular data to identify “opportunity gaps” that should be addressed to increase program effectiveness. Having robust participation of different stakeholders in the creation and administration of the M&E system, including agreed upon indicators, targets and objectives, ensured broad commitment to, and use of the data.[3]
2.6 “Know Your Epidemic” District Profiles and Action Plans
The scale up of the HIV program to cover the whole country in NACO 3 required a massive expansion of data systems and research, and more data suddenly became available. In 2010, districts were required to triangulate this data by writing a “know your epidemic” district profile. This ensured the data was analyzed and then used for planning as part of the creation of an annual action plan. This was first done as group work according to a set format with clear pointers, asking teams to identify external and internal risk factors, who the risk populations were, where they were located, and if program efforts were reaching them. Teams also had to describe program components and if they were suitable for the identified needs. Lining up the needs alongside program components was an especially effective component of planning.
Templates were also provided for the action planning and budgeting processes to ensure all plans and budgets were evidence based. This ensured data use was integrated into the regular program cycle. This occurred across 540 priority districts in the country.
2.7 National Data Analysis Plan
During the fourth phase of NACO (NACO 4), the key strategies include “strengthening Strategic Information Management Systems” and one of the important guiding principles was “evidence based and result oriented program implementation”. Towards these strategies, under the National Data Analysis Plan, all of the routine data sets from NACO were made available to researchers to establish more definitively what works in turning back HIV. Public health institutes worked with program managers to come up with priority research questions. This took 3-4 months of brainstorming and refinement to identify which questions could be best answered by secondary data. This was followed by data analysis workshops and scientific writing workshops. Out of 40 topics, 32 progressed to scientific writing, and about 16 were accepted for publication with 5 more nearing completion.[4]
The program managers required hand-holding in terms of data analysis but the researchers also needed support in understanding which research questions had practical salience. The idea of a journal authorship made program managers excited to look at the data, and have questions related to their work answered. Furthermore, the partnerships meant an expansion in the knowledge base of what works in combatting HIV in India.
3. Innovative Data Applications: Categorization of Districts for Targeted Programming
One example of an innovative data-driven strategy is the categorization of districts for differentiated programming. Prior to NACP III, resources would be distributed according to where there was a good NGO presence, not necessarily according to need. Evidence about prevalence and the localized nature of the epidemic helped even out such distortions. Given the heterogeneity of the HIV epidemic in India, districts were categorized for differential programming among subpopulation groups. Indicators used included ANC HIV prevalence (a proxy for the prevalence in the general population), and high-risk group (HRG) prevalence. High priority districts were given increased resources, more policy attention and stronger managerial capacity. In these priority districts, District AIDS Prevention and Control Units (DAPCU) were also established to ensure better convergence of services and closer monitoring. Differentiated planning was a unique feature of NACO, it helped more efficient use of scant resources, by providing a clear focus on local needs.
Key Lessons Learned
A review of the experience of NACO, the Technical Assistance Units and program implementors provides useful lessons for public health programs in fostering data use. These are described below.
Leadership and a Systems Approach: The successes of the HIV program highlight the importance of leadership and having human resources in place. This includes an M&E division, and an epidemiologist so different types of data can be reviewed, and in the case of the State AIDS Control Societies, a Project Director. Once you have the right leadership, a systems approach for improving data use is possible – this means integrating data use into routine program processes such as planning, supervision, monitoring and implementation.
Dedicating Time for Data Review in Meetings: Teams typically need to be asked to review data at meeting platforms. One or two hours needs to be set aside for data review and discussion. If time is not dedicated for this activity, it is likely to be skipped.
Supervisory Visits: All supervisory visits need to pay attention to data registers and records, at all points in the data value chain.
External Partners: An external partner to analyze the data for government decision makers was important in some administrative contexts where the demand for data was not high, and the quality of data poor. This role was performed by the TSUs in some states and the national level. The key to ensuring data use was the external partner presenting it in the right way – it should never look as though its critiquing the program, only helping to improve it.
Quality: One of the biggest challenges is data quality and if this is not addressed efficiently, efforts to promote data use will lose momentum quickly.
Acknowledgements
This case study was put together by Anna Schurmann, Manu Panjikaran, Vishal Shastri and Girdhari Bora. Thank you to Yujwal Raj Pinnanameni (former head of Strategic Information at NACO) and Shajy Isac (Executive Director, IHAT) for describing their experiences and strategies.
[1] Sgaier, S.; Anthony, J.; Bhattacharya, P.; Baer, J.; Malve, V.; Bhalla, A.; Hugar, V. Strengthening Government Management Capacity to Scale Up HIV Prevention Programs through the Use of Technical Support Units: Lessons from Karnataka State, India. Global Health: Science and Practice. November 25, 2014 as doi: 10.9745/GHSP-D-14-00141 www.ghspjournal.org [2] This was primarily due to surveillance sites being expanded from high-prevalence urban areas, to the whole country and including low prevalence rural areas, and therefore fewer pregnant women testing positive for HIV. [3] Thompson, L. et al. 2012. A Systematic Approach to the Design and Scale-up of Targeted Interventions for HIV Prevention Among Urban Female Sex Workers. Karnataka Health Promotion Trust. Bangalore. [4] National AIDS Control Organisation (NACO). 2015. A Report on National Data Analysis Plan: A Systematic Approach Towards Analysis of National HIV/AIDS Programme Data. New Delhi: NACO, Ministry of Health and Family Welfare, Government of India. [1] Blanchard, J.; Bhattacharya, P.; Kumaran, S.; Ramesh, B.M.; Kumar, N.S.; Washington, R.G.; Moses, S. 2010. “Concepts and Strategies for Scaling Up Focused Prevention for Sex Workers in India”. Sex Transm Infect 2008;84 (Suppl II):ii19–ii23. doi:10.1136/sti.2008.033134 [2] Thompson, L. et al. 2012. A Systematic Approach to the Design and Scale-up of Targeted Interventions for HIV Prevention Among Urban Female Sex Workers. Karnataka Health Promotion Trust. Bangalore. [3] Thompson, L. et al. 2012. A Systematic Approach to the Design and Scale-up of Targeted Interventions for HIV Prevention Among Urban Female Sex Workers. Karnataka Health Promotion Trust. Bangalore. [4] Blanchard, J.; Bhattacharya, P.; Kumaran, S.; Ramesh, B.M.; Kumar, N.S.; Washington, R.G.; Moses, S. 2010. “Concepts and Strategies for Scaling Up Focused Prevention for Sex Workers in India”. Sex Tran sm Infect 2008;84 (Suppl II):ii19–ii23. doi:10.1136/sti.2008.033134 [5] Sgaier, S.; Anthony, J.; Bhattacharya, P.; Baer, J.; Malve, V.; Bhalla, A.; Hugar, V. Strengthening Government Management Capacity to Scale Up HIV Prevention Programs through the Use of Technical Support Units: Lessons from Karnataka State, India. Global Health: Science and Practice. November 25, 2014 as doi: 10.9745/GHSP-D-14-00141 www.ghspjournal.org [6] Avahan, 2008. Use It or Lose It: How Avahan Used Data to Shape Its HIV Prevention Efforts in India. Bill & Melinda Gates Foundation. New Delhi, India. [7] Thompson, L. et al. 2012. A Systematic Approach to the Design and Scale-up of Targeted Interventions for HIV Prevention Among Urban Female Sex Workers. Karnataka Health Promotion Trust. Bangalore.
[1] After South Africa and Nigeria. UNAIDS, 2017. Data Book. Available online at: https://www.unaids.org/sites/default/files/media_asset/20170720_Data_book_2017_en.pdf. Accessed on 10th September 2019 [2] National AIDS Control Organization, National Institute of Medical Statistics, 2012. Technical Report: HIV Estimates 2012. Available online at: http://files.unaids.org/en/media/unaids/contentassets/documents/data-and-analysis/tools/spectrum/India2012report.pdf. Accessed 9th September 2019. [3] Rao, S. 2017. Do We Care: India’s Health System. Oxford University Press, New Delhi.