Nishtha Bharti and Naveen Thayyil
As the Indian government eagerly sets out to envision a ‘futuristic digital health system that would undoubtedly be the very best in the world’, it has managed to circumvent a key variable which is crucial to the promotion of technological interventions as the cure to India’s crumbling healthcare system. This variable pertains to a reflexive assessment of the present moment – a prerequisite for this ambitious transition. Various milestones in this journey – the National Health Policy (2017), the Digital Information Security in Healthcare Act (2017), the National Health Stack (2018), the National Digital Health Blueprint (2019) and most recently, the National Digital Health Mission (2020) – collectively propose the architecture and protocols for electronic health data registries, information exchanges, data interpretability , a unique health identifier and a health analytics platform. In short, they put forward slivers and templates for data governance in service of a healthcare information technology approach, but without a substructure to support this edifice. This substructure involves the availability of adequate health data – a feature that itself suffers from disparities. This paper will argue that such disparities in health datasets can be traced to the existing state of healthcare inequities in India. And while the digitisation drive seeks to rectify these inequities, it cannot possibly do so without circling back to their fundamental nature and cause.
The paper will further explore what is missing, by design, in the above stated documents: a discussion of what gets considered as health information and an accountability for the limited datasets that characterise the health information infrastructure in India. With specific reference to an emerging technology that is grounded in data – Artificial Intelligence (“AI”), this paper will posit that policy-makers in India are engaged in shaping the production of knowledge – and correspondingly, ignorance – through a deliberate, purposeful move. This involves, in first instance, bypassing the discussion regarding the availability of adequate medical data and its pertinence to the reliability of viable AI models. And simultaneously, taking the leap to elaborate how data ought to be governed so as to advance digital health initiatives.
The next section will detail how digital futures have been envisioned in recent health policies through data-driven health management systems. Section three will underscore the inadequacy of existing health data in India and its misalignment with the notion that digitisation of health data and AI interventions will improve access and outcomes. Section four will mobilise the literature on the politics of knowledge-making and strategic ignorance to demonstrate how the dilemmas inherent in decreeing data-driven, information-premised futures – that warrant critical societal attention – are eluded.
II. DIGITAL FUTURES AND HEALTH POLICIES
The Indian government’s commitment to a shift towards data-driven health management and creation of digital health information architecture is apparent in the National Health Policy (2017), which envisages ‘extensive deployment of digital tools for improving the efficiency and outcome of the healthcare system’. It stipulates the setting up of a National Digital Health Authority (“NDHA”) to oversee digital health initiatives as well as the creation of Health Information Exchange platforms and Patient registries to facilitate big data analytics. These data networks, the policy suggests, can be developed in alignment with Metadata and Data Standards (“MDDS”) and Electronic Health Record (“EHR”), and further, potentially be linked to the national Unique Identification Database, Aadhaar. This layout of a digital health technology ecosystem is elaborated in the National Health Stack: Strategy and Approach (2018) (“NHS”) which proposes to bring into a play ‘a powerful technology arsenal’ to effectuate ‘a complete redesign of the flow of people, money, and information, as well as a layered approach to providing comprehensive foundational health functions (…)’. Within its schematic, the NHS proposes the following components: National Health Electronic Registries, a Coverage and Claims platform, a Federated Personal Health Records Framework and a National Health Analytics Platform. These modules, in addition to Digital Health IDs and Health Data Dictionaries, are expected to overhaul the health information infrastructure of India, with the hope of steering ‘smart policy making, for instance, through improved predictive analytics’. Significantly, a NITI Aayog publication (2019) resulting from a workshop titled Health System for a New India: Building Blocks ascribes the ideation of the NHS to ‘success of the previous Aadhaar initiative’ and reiterates that the NHS will open the door for comprehensive healthcare data collection across the country.
The implementation framework for the NHS was proposed in the National Digital Health Blueprint (2019) (“NDHB”), by a Committee constituted under the aegis of the Ministry of Health and Family Welfare. The NDHB details the ‘building blocks’ and institutional mechanisms for data integration of the health system in India. This involves instituting Health Master Directories & Registries, Data hubs and Personal Health Identifiers (relying partly on Aadhaar-based Identification/ Authentication) alongside a Digital Health India portal and MyHealth App. These provisions are expected to mediate the generation, collection, exchange and standardisation of health data across various channels and to eventually advance health data analytics and decision-support systems. The NDHB also recommends the establishment of a specialised organisation, the National Digital Health Mission (later renamed as ‘Ayushman Bharat Digital Mission’) – which was launched in August 2020. The Ayushman Bharat Digital Mission (“ABDM”) essentially seeks to advance the evolution of the National Digital Health Ecosystem, by creating a centralised database to store the details of a unique health identifier for every citizen, a Digi-Doctor, a repository of healthcare facilities across the country and digitising patient health records in the form of Electronic Medical Records.
The stated purpose of fashioning these information systems is for enabling interoperability and portability of health data and accelerating the subsequent provision of health services. Furthermore, the strategy paper of the ABDM identifies the ‘current strong public digital infrastructure, including that related to Aadhaar, Unified Payments Interface and wide reach of the Internet and mobile phones (Jan Dhan-Aadhaar-Mobile trinity)’ as a propitious asset for setting up the building blocks of ABDM. Together, these policy instruments configure a certain vision of information governance, which does indeed conceptualise regulation of a certain kind – proper interfaces and protocols for data storage, processing, sharing, verification etc. When read alongside the Digital Information Security in Healthcare Act (2018), there is also a discernible concern regarding the nature of ownership of patient health data, confidentiality of personally identifiable information, privacy and consent frameworks. But what still remains unaddressed, as will be elaborated in the next section, is the fact that existing health data in India, which is the object of the digitisation fantasy, is fragmented and partial. Any AI ‘solution’ leveraged from such databases will only replicate the inequitable materiality of the present healthcare system, instead of improving access and outcomes – the often cited assurance on which AI interventions in the healthcare domain are pitched.
III. BIG DATA, AI APPLICATIONS AND HEALTH REALITIES IN INDIA
The deployment of AI in the healthcare domain and its research uptake in India can be categorised across four broad verticals. First, applications that are anchored in organising medical information, hospital/practice inventory management, patient admission, discharge and post discharge monitoring, managing operations for insurers – in general, those aiming at improving the workflow of health systems. Second, there are mobile health apps and wearables, mostly offering real-time lifestyle tracking and wellness analytics. These also include health lab and ambulance aggregators as well as chatbots. Third, AI-enabled disease surveillance systems are deployed for prediction of mass infections and early detection of disease outbreaks. Finally, AI models are employed in the realm of predictive diagnostics, such as image stratification and interpretation across various radiology and pathology tests for diagnostic screening and assistance. Identification of patterns is the major driver for such algorithms, mostly utilizing deep neural networks. A key variable on which the robust development of these applications critically rests is appropriate health data.
There are two broad difficulties in the approach to a data-driven transformation of the healthcare system in India – a trajectory that the Indian government seems to have decisively committed itself to, in its eagerness to make the most of AI ‘solutions’.
First, big data in the healthcare sector is primarily sourced from Electronic Health Records (“EHR”s), laboratory reports, pharmacy prescriptions, insurance and other administrative logs, as well as real time metrics collected from sensor-based devices, wearables and health applications. Additionally, the digital footprint left by consumers on social media – such as blogs, posts and searches – is also being leveraged for scanning and assessing health risks. It is crucial to note that all these can only be linked to individuals who have sought healthcare services at some point or those who own and operate devices provisioned with digital health assistants or those who are regular users of social media and have the resources (internet enabled handsets/desktops) to sustain a consistent engagement with such platforms. It does not include those persons who are factored out of the big data trail because of inability to access fundamental healthcare services or integrate into the ecosystem of digital health platforms or afford the time and amenities to become active social media users. These excluded groups bear the brunt of the dismal state of the healthcare system in India and whose plight the above mentioned policies are purported to address, even as these policy instruments ignore the fact that the solution endorsed itself builds on existing health disparities.
As has been well-documented in the literature on Critical Data Studies, such injudicious reliance on data mining practices can replicate pre-existing patterns of exclusion and inequality experienced by historically disadvantaged and vulnerable groups. The problem, as succinctly captured by Jonas Lerman is this: ‘big data poses risks also to those persons who are not swallowed up by it – whose information is not regularly harvested, farmed, or mined’. It seems counter-intuitive then, to propose an elaborate information governance infrastructure to mitigate the disenfranchisement of those individuals who are not part of the initial information design in the first place.
Second, the datasets that are already available in the domain of health in India are fragmented, dispersed and incomplete. The National Health Profile (2019)  and the National Health Accounts: Estimates for India (2015-16)  both admit to limitations in the quality of primary health data collected to accurately reflect the reality of health variables at the national, sub-national and individual level. While the government collates large swathes of data, this data lacks precision, is error-prone and still excludes certain cohorts who fall outside the ambit of mass sample studies. Even if this longitudinal data is somehow successfully digitised and made interoperable, it is likely to be unrepresentative along various parameters – not a desirable preamble to build reliable AI models. The present regulatory paradigm in digital health initiatives is focused on the assembly and standardisation of health records, their oversight and exchange. But the issuance of directions along these lines, no matter how well-meaning, does not take away substance from the fact that adequate health data, the foundational building block of this approach, is still not a given in India. And this existential difficulty circles back to the original problem of the dire state of India’s healthcare system, to mitigate which the entire project of data-driven health management and information architecture is routinely advanced.
The healthcare system in India is presently mired in a number of difficulties, an acute shortage of medical professionals, precarious standard-of-care in most public establishments, an extractive and unreliable private sector, continual neglect of preventive and community health measures, and high reliance on out-of-pocket expenditure. Concurrently, India’s public expenditure on health remains one of the lowest in the world, a little over 1% of GDP for more than a decade. But this receives little scrutiny in all the policy pronouncements and strategy statements relating to futuristic imaginations of the health ecosystem. Instead, the Indian government has abruptly spring-boarded to proclaiming an allegiance to the latest digital technologies and data analytics (such as through the various policy pronouncements enumerated in the previous section), and to prescribing the paraphernalia that it deems necessary to actualise its vision of a datafied, digital health ecosystem. An instantiation of the latter is the NITI Aayog Working Document titled Towards Responsible #AIforAll released in July 2020, which delineates the system and societal considerations in the usage of AI, the principles for ensuring its responsible development and deployment, and related enforcement mechanisms to streamline its uptake in various sectors. In the same vein, the National Digital Health Mission: Health Data Management Policy, also approved in 2020, provides the standards for safeguarding digital personal health data within the ambit of the National Digital Health Ecosystem (“NDHE”). Such initiatives can be seen to be at par with the global thrust to enunciate accountability frameworks for emerging technologies and the data regimes that they precipitate. At the same time, they can also be read as an act of bypassing the very possibility of a dialogue regarding the need for sophisticated technological interventions at all and the more pressing necessity for more basic, foundational healthcare reforms. In a way, the proclivity to indulge in detailing the accountability mechanisms for information structures and AI models obfuscates the prospect of holding the state accountable for a decent health system. Accordingly, the next section asks the question – what specific manoeuvres of knowledge-making orient technological decisions and normalise manufactured ignorance?
IV. MANUFACTURED IGNORANCE
Many studies on the politics of knowledge-making are predicated on the realisation that the shared willingness of individuals to subscribe to ignorance is fundamental to expert control and authoritative expansion. In his essay ‘The Ecology of Ignorance’ (1998), German sociologist Niklas Luhmann noted that the purposeful cultivation of ignorance clears the way for justifying non-liability for one’s actions.  He points out that those having political or cultural power are well aware that a commitment to knowing implies that they can potentially be held responsible for their knowledge whereas posturing ignorance provides a reprieve from having to answer for future consequences. In the Indian context, the persistent focus on a data-driven, digitised, AI-based health ecosystem as described in the sections above serves to spotlight a particular approach to transforming healthcare whose desirability is framed to the exclusion of alternative approaches. To begin with, this singular focus is complemented with the strategy of evading any discussion about the limitations of India’s health information infrastructure, which is the bedrock of the proposed approach. The absence of such a discussion means that concerns regarding the socio-ethical implications of limited health datasets – upon which AI models are being envisioned – are circumvented
Notably, such a move also eludes an acknowledgment of the idea that the propagation of advanced digital technologies in healthcare potentially advantage certain groups/interests over others as well as any meaningful dialogue about the government’s responsibility for such consequences. This strategic enterprise of advancing certain policy agendas, while deliberately maintaining ignorance about specific issues that render those agendas questionable, unfolds on various levels. This section attempts to situate such a strategy in the literature that identifies ignorance as an epistemological asset.
Ignorance, as described by Luhmann (1998) serves important political and social functions, wherein an admission of not knowing and not understanding implicitly absolves someone from assuming any blame. This assertion has been probed thoroughly in recent years by an increasing number of scholars. For instance, Frickel and Vincent (2007), in their analysis of the aftermath of Hurricane Katrina in the US and expert understandings of contamination in New Orleans highlight ‘organized ignorance’ as an effective tool that diminishes the socio-historical contexts in which regulatory knowledge is created. They contend that as a consequence of this designed ignorance and the subsequent shift in public discourse, expert systems and regulatory agencies are spared from answering some of the ‘most basic questions concerning safety, health, and sustainability’ of a disciplinary field.
This insight is distinctly reflected in the policy documents and position papers of the Indian government, which favour overarching narratives around precision medicine while deflecting difficult questions regarding the representational value of health data in the country, and the causal factors behind that shortcoming. This knowledge-making practice that is essentially built on ‘organized ignorance’ gives legitimacy to the fundamental leap of faith in championing big data analytics and AI in healthcare, even as rudimentary healthcare services and infrastructure in India are not guaranteed to citizens.
Linsey McGoey (2007, 2012) has further explored the benefits that accrue to actors by maintaining strategic ignorance. She points to a ‘will to ignorance’ (2007) that acts as a valuable organisational resource, especially for regulatory bureaucracies. Focusing on pharmaceutical drug regulation, she argues that any perceived complexities that may arise in the face of unpredictable outcomes can be easily avoided by strategic ignorance. She observes that ‘the deliberate refusal to become informed of conclusive evidence that might be counterproductive to organizational or individual aims’ (2012) allows individuals and institutions to deny liability in crisis situations. Ignorance, then, becomes an asset and is operationalised through various tactics.
In the Indian context, reliance on strategic ignorance also allows the policy-makers to disavow themselves of the knowledge that the overenthusiastic push for digital information structures will eventually serve certain interests more than others. Specifically, it camouflages the inevitability of specific market responses that these policies are set to trigger – in particular, the entry of private players in not only delivering the professed vision of ‘revolutionising’ healthcare, but also contributing to developing that vision alongside the government. A preview to this scenario is the commissioning of iSpirt, a domestic industrial guild of private software developers, to design the infrastructure for the National Health Stack, based on the tech stack of Aadhaar. The leveraging power afforded to the private sector through such assignments – in the face of prevailing inadequacies in access, service delivery and expenditure in the healthcare domain – undermines the egalitarian tenor of the policy pronouncements of the Indian government. And as such, this outcome does not find mention in government documents, which are simply overrun with optimistic accounts of digitisation and its governance.
Another valuable anchor to understanding knowledge-making practices is offered by Kleinman & Suryanarayanan , who have noted that through the lens of ‘epistemic form’, individuals in academic, regulatory, and corporate organisations institutionalise various types of ignorance by privileging a ‘suite of concepts, methods, measures, and interpretations’. Such a tendency, they argue while examining the Colony Collapse Disorder (CCD) in the US, influences both research agenda setting and regulatory policies. As evident in the policy tracts of the Indian government, certain frameworks of theoretical assumptions as well as standards of evidence that look towards certain directions, but leave unexamined other ways of understanding, serve to enhance knowledge gaps that could have reasonably catered to ethical concerns regarding AI applications in healthcare.
One such instantiation is in the Government of India extolling the virtues of Aadhaar and building on its perceived successes to advance the agenda of digitisation of health. It has been well documented how the state strategy for improving service delivery through the technology of digitisation and the seeding of Aadhaar with various services resulted in further hurdles of inaccessibility for the vulnerable sections of the society, thereby creating novel paradigms of inclusion and exclusion. Yet, various proposals on the future of healthcare in India visualise Aadhaar as an attractive reference point to emphasise the transformative potential accruing from its inherent virtue. To borrow from Kleinman & Suryanarayanan, this allegiance to the ‘epistemic form’ of a techno-deterministic logic leaves unexamined other ways of approaching the project of improving the healthcare situation in India.
It is precisely these kinds of knowledge gaps and the methodical entrenching of ignorance towards the complexities of information systems that affords the policy-makers in India a leeway to introduce authoritative imaginations of digital health, even as they side-step any discussion regarding the desirability and preparedness of this ambitious overhaul of the healthcare system in India. This proclivity also shapes the manner in which the government imagines and executes its responsibilities, and allows it to disavow itself from any accountability for the present state of the healthcare system in India. The ‘policy silences’ about questions that are left unasked, decisions that are not made and values that are not prioritised in public debates are all conducive to the production of ignorance, and to the dismissal of existing difficulties in assembling a robust data-driven digital health information architecture.
The central thesis for this paper is that digital information initiatives introduced with the promise of improving the healthcare system in India will be undermined by the existing problems of that very system. Those problems prohibit the inclusion of certain cohorts into the composition of initial health data, thereby transferring their disadvantageous circumstances onto the new digital data infrastructures that are currently being envisioned. Such an assertion, however, does not presume that inclusion in health information systems is seamlessly congruous with enjoyment of one’s health rights and availing health services. But instead, exclusion from those systems is worth being investigated not only to assess the inherent promise of the vision being endorsed through the latest digital technologies, but also to reveal the latent (non) decisions that lie in the shadow of that promise.
The fact that the National Digital Health Mission (2020) was announced in the midst of a public-health crisis engendered by a raging pandemic which painfully brought to fore the fragile state of the healthcare infrastructure in India, underscores the notion that any scruples about addressing or even acknowledging the gaping lacunae in the healthcare system have long been set aside. Instead, as this paper has sought to demonstrate, the politics of ‘ignorance by design’ allows the government to get away with informational inadequacies in the digital data systems it propagates. In particular, this form of knowledge-making occludes the reality of the fragile foundations of the Indian healthcare system, its limitations and the consequences of those limitations on the desirability of technological interventions. In the haste to advocate the potential of data-driven, algorithmic promises, questions regarding the configuration of health information and its potential for inclusivity are systematically unasked. What remains is a scaffolding of a programmatic vision that simply skims the surface of what a transition towards digital health entails in reality.
- NITI Aayog, GoI, National Health Stack (NHS) : Strategy and Approach (July 2018) 5 <http://niti.gov.in/writereaddata/files/document_publication/NHS-Strategy-and-Approach-Document-for-consultation.pdf> accessed 30 October 2020 ↵
- Ministry of Health & Family Welfare, GoI, National Health Policy, 2017 (2017) 25 ↵
- NITI Aayog, GoI, National Health Stack (NHS) : Strategy and Approach (July 2018) 6 <http://niti.gov.in/writereaddata/files/document_publication/NHS-Strategy-and-Approach-Document-for-consultation.pdf> accessed 30 October 2020 ↵
- ibid 12. ↵
- NITI Aayog, GoI, Health System for a New India: Building Blocks (November 2019) 247 <https://niti.gov.in/sites/default/files/2019-11/NitiAayogBook_compressed_1.pdf> accessed 3 November 2020 ↵
- National Health Authority, Ministry of Health & Family Welfare, GoI. National Digital Health Mission: Strategy Overview, (July 2020) <https://ndhm.gov.in/documents/ndhm_strategy_overview> accessed 15 November 2020 ↵
- DISHA has now been subsumed under the ‘Data Protection Framework on Digital Information Privacy, Security & Confidentiality’ Act. See <https://pib.gov.in/Pressreleaseshare.aspx?PRID=1578929> accessed 10 December 2021 ↵
- Ministry of Health & Family Welfare, GoI, Digital Information Security in Healthcare Act [Draft for Public Consultation] (November 2017) <https://www.nhp.gov.in/NHPfiles/R_4179_1521627488625_0.pdf> accessed 7 November 2020 ↵
- Smadar Shilo, Hagai Rossman and Eran Segal, ‘Axes of a Revolution: Challenges and Promises of Big Data in Healthcare’ (2020) 26 Nature Medicine 29 <https://doi.org/10.1038/s41591-019-0727-5> ↵
- Rakesh Raja, Indrajit Mukherjee and Bikash Kanti Sarkar, ‘A Systematic Review of Healthcare Big Data’ (2020) Scientific Programming 1< https://doi.org/10.1155/2020/5471849> accessed 19 December 2020 ↵
- Susan L. Erikson, ‘Cell Phones ≠ Self and Other Problems with Big Data Detection and Containment during Epidemics’ (2018) 32(3) Medical Anthropology Quarterly 315; Sundeep Sahay and Arunima Mukherjee, ‘Where Is All Our Health Data Going?’ (2020) 55(1) Economic & Political Weekly 46 ↵
- Kate Crawford, ‘The Hidden Biases in Big Data’ Harvard Business Review (Apr. 1, 2013) <https://hbr.org/2013/04/the-hidden-biases-in-big-data> accessed 21 January 2021; Solon Barocas and Andrew D. Selbst, ‘Big Data’s Disparate Impact’ (2016) 104 (3) California Law Review 671; Sarah E Malanga and others, ‘Who’s Left Out of Big Data? : How Big Data Collection, Analysis, and Use Neglect Populations Most in Need of Medical and Public Health Research and Interventions’ in I. Glenn Cohen and others (eds.), Big Data, Health Law, and Bioethics (Cambridge University Press 2018). ↵
- Jonas Lerman, ‘Big Data and Its Exclusions’ (2013) 66 Stanford Law Review Online 56 <www.stanfordlawreview.org/online/privacy-and-big-data-big-data-and-its-exclusions/> accessed 4 December 2020 ↵
- Central Bureau of Health Intelligence, Ministry of Health & Family Welfare, GoI, National Health Profile 2019 (October 2019) <http://www.cbhidghs.nic.in/WriteReadData/l892s/8603321691572511495.pdf> accessed 25 November 2020 ↵
- Ministry of Health & Family Welfare, GoI, National Health Accounts Estimates for India (Financial Year 2015-16) (2018) <https://main.mohfw.gov.in/sites/default/files/NHA_Estimates_Report_2015-16_0.pdf> accessed 25 November 2020 ↵
- Rakhi Dandona, Anamika Pandey and Lalit Dandona, ‘A Review of National Health Surveys in India’ (2016) 94(4) Bulletin of the World Health Organization 286; Rajeev Kamal Kumar, ‘Quality Issues in the Health Management Information System: A Case Study of Bihar’ (2018) 53(45) Economic & Political Weekly 22 ↵
- Sadhika Tiwari, ‘India Spent 1% Of GDP on Public Health for 15 years. Result is Vulnerability to Crises’ (IndiaSpend, 26 June 2020) <https://www.indiaspend.com/india-spent-1-of-gdp-on-public-health-for-15-years-result-is-vulnerability-to-crises/> accessed 19 November 2020 ↵
- NITI Aayog, GoI, Working Document : Towards Responsible #AIforAll (Part 1) (June 2020) <https://niti.gov.in/sites/default/files/2020-11/Towards_Responsible_AIforAll_Part1.pdf> accessed 16 January 2020 ↵
- National Health Authority, Ministry of Health & Family Welfare, GoI, National Digital Health Mission: Health Data Management Policy (2020) <https://ndhm.gov.in/documents/HealthDataManagementPolicy> accessed 29 December 2020. A revised draft of this policy was released in April 2022. See, Press Information Bureau, GoI, ‘National Health Authority releases revised draft of Heath Data Management (HDM) policy’ (April 23, 2022) <https://abdm.gov.in:8081/uploads/46926125ea897a0499717972f6bf1cfc_95c71293dc.pdf> accessed 29 May 2022 ↵
- Niklas Luhmann, ‘Ecology of ignorance’ in Niklas Luhmann (ed) Observations on Modernity (Stanford University Press 1998) ↵
- Scott Frickel and M. Bess Vincent, ‘Hurricane Katrina, Contamination, and the Unintended Organization of Ignorance’ (2007) 29 Technology in Society 181, 182 ↵
- Linsey McGoey, ‘On the Will to Ignorance in Bureaucracy’ (2007) 36(2) Economy and Society 212 ↵
- McGoey, ‘The Logic of Strategic Ignorance’ (2012) 63(3) The British Journal of Sociology 533, 562 ↵
- Nileena MS, ‘Is India Privatising Governance through Partnerships in Public Digital Infrastructure?’ (The Caravan, 20 October 2020) <https://caravanmagazine.in/policy/is-india-privatising-governance-through-partnerships-public-digital-infrastructure> accessed 2 July 2021 ↵
- NITI Aayog, GoI, National Strategy for Artificial Intelligence (June 2018) 20 <http://niti.gov.in/sites/default/files/2019-01/NationalStrategy-for-AI-Discussion-Paper.pdf> accessed 1 July 2021; NITI Aayog, GoI, National Health Stack (NHS) : Strategy and Approach (July 2018) 7 <http://niti.gov.in/writereaddata/files/document_publication/NHS-Strategy-and-Approach-Document-for-consultation.pdf> accessed 3 July 2021 ↵
- Daniel Lee Kleinman and Sainath Suryanarayanan, ‘Dying Bees and the Social Production of Ignorance’ (2013) 38(4) Science, Technology, & Human Values, 492 ↵
- Bidisha Chaudhuri and Lion König, ‘The Aadhaar Scheme: Cornerstone of a New Citizenship Regime in India’ (2017) 12 Contemporary South Asia1; Ritika Khera, ‘Impact of Aadhaar on Welfare Programmes’ (2017) 52(50) Economic & Political Weekly 61 ↵
- Harsh Vardhan, ‘Citizen’s Interest is at Core of NDHM’s Design. It will radically change India’s Health Delivery Landscape’ (The Indian Express, 29 August 2020) <https://indianexpress.com/article/opinion/columns/ndhm-health-id-covid-19-india-6574094/> accessed 1 July 2021 ↵
- Des Freedman, ‘Media Policy Silences: The Hidden Face of Communications Decision Making’ (2010) 15(3) The International Journal of Press/Politics 344 ↵