A Model for Priority Setting in Health Technology Innovation Policy

Health Technology Assessment focuses on the equal appraisal of health technologies introduced into the market. This has made regulators and the governance of innovation reactive and dependent on the initiatives that innovators take for technology development, thus making it supply-driven. The policy-makers’ role has become one of appraising technologies that are already developed rather than guiding the development agenda. This severely limits the possibility to ensure that health technologies sufficiently address major issues such as the burden of disease, trade deficit, and health inequalities. It places governments outside of the actor arena that co-shapes technologies in the early stages, restricting the involvement in facilitating whether to scale up or not. It makes it hard to achieve health technology governance practices that maximally contribute to ensuring technological developments that address public concerns. What is the potential of the framework for changing this dynamic and how can evidence shape technology development agendas without falling into the traps of regulator lock-in or social engineering? The methodology presented in this study takes the first important steps toward an evidence-based framework for priority setting to guide innovations, particularly in the health and social sectors.


INTRODUCTION
The appraisal of health technologies introduced in the market is of utmost relevance for healthcare governance. One of the greatest challenges that governments face is aligning the agenda of technology development with social indicators like the burden of disease and macro-economic indicators like trade. However, the role of evidence-based policy restructuring in guiding medical device development has remained an overlooked possibility. Understanding the necessities and gaps in medical device development could have major consequences for sectoral advancement and its benefits to the society instead of being locked in a supply-driven mode. As the current COVID-19 pandemic painfully shows, it is also necessary to take into account a country's capacity for self-sufficiency in terms of manufacturing the devices used in their territories and becoming more independent from importing medical devices and the economic impacts of such import. The pandemic shows that countries across the world, lower-middle-income countries (LMICs) as well as high-income countries (HICs), have all become largely dependent on the import of medical devices and that self-sufficiency by no means is a concern for emerging economies alone. This study aims to provide an evidence-based framework for priority setting in guiding innovations by developing a practical model that can be implemented using country-specific data that reflects the actual territorial needs and can be related to the countries' economic capabilities. A set of composite indicators have been identified and used for the priority setting exercise. Firstly, the quality of human life is a major indication of national economic progress and human development index. 1 The corollary of Gross Domestic Productivity as a human health welfare index is an indication of national health, 2 both economic and contextual. One major parameter for the identification of wellness of populace is public health data records. 3 All countries have their specific manner of maintaining public health records 4 and analysis methodologies 5 to use evidence from such records.
While minor variations in the analytical tools might be present, one common consensus among all government structured health analytics is the importance of mortality records. 6 Of the common parameters of assessing disease-affected livelihood, the Quality-Adjusted Life-Year is a generic measure of disease burden, including both the quality and the quantity of life lived. The Institute of Health Metrics and Evaluation (IHME) at the University of Washington published a report titled WHO Global Burden of Disease (GBD) which considered Disability-Adjusted-Life-Year (DALY) as the point of consideration for disease-affected livelihood. 7,8 To consolidate the data of the major disease burden for India, this study focused on interpolating the diseases that are of immense concern. Each disease has a large number of diagnostic, therapeutic, rehabilitative, and palliative procedures meant for combating the condition. While the diseases are classified under the International Classification of Disease (ICD-11) by WHO, the interventions to tackle such diseases are listed under the International Classification of Health Interventions also by WHO. The interventions are thereafter dependent on various health technologies, and while pharmaceutical products often come with varied alternatives, most medical devices do not have an alternative for the patient nor operationally for the care provider. For example, while several lines of drug therapy exist for non-communicable diseases such as renal failure or diabetes, there are no alternatives to a dialysis machine, a dialyzer, a glucometer, or an insulin pump. There lies an important distinction between medical devices, with both devices and drugs being health technologies.
Furthermore, by conceptually combining the evidence from the burden of disease estimates for a country to the export-import trade data on medical devices, a model list could be enumerated to estimate which medical devices could be developed in a country, which would be reflective of its health as well as economic impact. This reasoning was applied in this study to India's context. As a country, India currently imports >80% of its medical device needs. 9 This ranges across all healthcare paradigms as well as all domains of devices. 10 It greatly affects the healthcare cost attributed primarily to capital expenditure on commodities as essential as health technologies. 11 The overall medical devices market in India is estimated to be USD 7 billion, 12 however the country imports over 80% of its needs, making medical technology acquisition costlier which negatively impacts the healthcare costs. The medical device market in India is growing at a 15.8% CAGR (compound annual growth rate) and is postulated as the fourth largest potential globally.
The methodology described in this paper, using a model that could cross-pollinate information from both disease burden and trade deficit has several advantages. Firstly, avoidable death or disabilities from disease or ailments that could be possibly cured by medical technology is an effective indicator of meaningful technological and scientific progress. Secondly, the import dependency on such devices including life-saving ones has a direct impact on the country's trade deficit impacting its macro-economic growth and therefore societal progress. Needless to say, United Nations in its 17 Sustainable Development Goals has given equal importance to these by keeping poverty elimination as Goal 1; Good Health and well-being as Goal 3; and Industry, Innovation, and Infrastructure as Goal 9.
Aligning these goals for priority setting in medical technology innovation could, therefore, result in health improvement and economic sustainability. This would also help realign the trade and export-import decision-making processes to encourage domestic manufacturing, which is increasingly important in light of world-wide outbreaks of infectious diseases. To the best of our knowledge, this study presents the first-ever attempt in correlating these essential principles to establish a pathway for medical technology development and drive innovation policy to improve healthcare access, economic sustainability, and societal impact.

METHODS AND RESULTS
In this section, the creation of a model or methodology to enumerate the priority list of medical devices that need to be developed is explained. The methodology is also a part of the results themselves, given that this study focuses, first, on the creation of a model and then application of the model to do the priority listing. For this reason, the authors opted to merge the methods and the results in the same section.
To estimate disease burden, the main causes of mortality compared over a decade for India, estimated by the IHME was used and are tabulated below (see Table 1). Trends from 2005 to 2015 signify that while there has been a substantial decline in mortality due to neonatal complications, mortality due to metabolic disease and/ or lifestyle diseases has overtaken the mortality due to communicable diseases -a classic trend across developing economies. Table 2 signifies similar parameters in the disability or morbidity estimates since mortality can be indicative but does not singularly affect GDP or macro-economic progress in a population.
The aspects under study were disease models which required significant medical device intervention, hospitalization, or otherwise. To outline the relevant causative factors and disease conditions, the major disease burdens were classified as communicable and non-communicable. The top five from both were enlisted and relevant treatment procedures that required technological interventions were detailed.
Concomitantly each of the medical devices used for diagnosis and treatment were mapped for each of the therapies corresponding to the diseases. Further, the segmentation of medical devices was charted by bringing in common devices used for these diseases. This had a relationship of one-to-many (e.g., cardiac and pulmonary diseases require more than one medical device) but also of many-to-one (hollow fiber membrane finding application in dialysis, oxygenator, ECMO and the like) ( Figure 1). Table 3 entails the communicable diseases list and technological interventions required to tackle such disease conditions. Table 4 enlists the non-communicable diseases and their diagnostic methods involving technological  interventions that are critical to disease treatment or mitigation.  Tables 1 and 2). (2) Identify diagnosis methods and device-related interventions for the top 5 diseases (see Tables 3 and 4). (3) Create a consolidated list of essential medical technologies (see Table 7) by combining the priority lists of medical and diagnostic devices for communicable and non-communicable diseases (Tables 5  and 6). Next, starting from the right: (4) Tabulate export and import data of categorized medical devices (see Table 8). (5) Consolidate high import and low export-dependent devices (see Table 9) as they require the increase or improvement in internal manufacturing capability to create self-dependency, higher affordability, and greater access. Data were collected by taking into account the top five communicable and top five non-communicable diseases which account for the most lives lost as per the latest data available.
With the perspectives of disease paradigm enlisted on priority, our methodology brought in the next component of these diseases and their possible intervention to prevent prevalence or provide treatment. Matching was done between diseases and relevant medical devices of the relevant intervention procedure from the prior lists removing overlapping entities if any. Table 5 lists the priority medical and diagnostic devices in communicable diseases subset, while Table 6 details the priority medical and diagnostic devices list for non-communicable ones. By merging the two previously referred lists and removing overlapping entities, a common list of essential medical technologies was created as shown in Table 7, consolidating this entire dataset. This list is referred to as the "priority list of medical and diagnostic devices on the  A second dataset was also created, referred to as the priority list in medical devices from the perspective of trade using export-import data as available in the public   domain. This list looked at medical devices apropos their HS codes. The Harmonized Commodity Description and Coding System, also known as the Harmonized System (HS) of tariff nomenclature is an internationally standardized system of names and numbers to classify traded products. The entire export and import data of categorized medical devices were tabulated as per recent import figures compared against the past few years of export (Table 8). It was postulated that threshold or higher exports indicated self-dependency, higher affordability, and greater access. By opposition, the extreme or high import dependency denoted greater costs and lower accessibility. Figure 2 illustrates the two types of scenarios mentioned, A and B, respectively. The consolidated high import and low export-dependent devices (driven from the illustrative scenario B) were then highlighted and tabulated into segments that required an increase or improvement in internal manufacturing capability. Table 9 forms the other half of the dataset of the study from the trade perspective.
Further, two of the aforesaid priority lists (Table 5 and  Table 9) were overlapped in a Venn diagram format, creating an intersection area of the priority list. The outcome of this entire exercise (Figure 3) was then subjected to expert discussion. The expert group included public health experts, epidemiologists, academia, research scientists, and user specialists.
Using the listed medical devices in the priority list that qualified expert group approval, these medical devices were mapped to their current domestic manufacturing

Burden of Disease
Import Dependency National Priority List capability. This was named as the Priority List of Health Technology and was used as a key approval criterion by a public agency for providing financial support for further research.
As some devices in the list could have domestic production viability for some of its components but would be completely dependent on exports for other components, it was also critical to understand which critical parts of components needed additional or focused research. To understand the key components of these technologies, two-day technology consultation was organized. This was named as "FIRST" (Formative Industry Leaders Research Institutes Start-up Partners Technology Meet) to identify essential components for focused research. The technology consultation included innovators, researchers, academia and industry, in which the shortcomings were highlighted and the technology development pathway discussed.
As part of the consultative process, a list of 108 Core Technology Components was identified. The list was then submitted to the concerned agencies within the government that provide funding for technology research. Requests for proposal for these were subsequently released by the 3. Health Technologies that focus on disease/clinical conditions reflective of LMICs/specific geographies and whose research would not be priorities in any other geography -such as snake bites.

DISCUSSION
This study aimed to provide an evidence-based framework for priority setting to guide innovation in the healthcare sector. By analyzing a combined dataset of the top ten diseases that account for the most lives lost and trade impact in India, this study specifically developed a method to determine a priority list of medical devices to address both the rising burden of diseases and growing trade deficit in the medical technology sector. Traditionally, the role of policy-makers has been to focus mainly on appraising technologies that are already selected by the innovators for research due to knowledge or engineering capabilities available with the innovators. Therefore, the policy maker's role has been reactive and necessarily not reflective of actual healthcare needs. This methodology allows the policy maker's approach to shifting from being merely reactive to actively driving the agenda of technology development. To do this, it is necessary to find, based on data, what the sectoral needs for development are. To the best of our knowledge, this study is the first to develop a transparent and intuitive method to establish the National Priority List of Medical Devices Research. The results of this study are twofold. Firstly, the formulation of the model can be considered a result in itself. Secondly, the medical device list that resulted from the application of the model/method developed for India is another result.
For the first set of findings, such as the method/model creation, it is important to consider that given its intuitive steps it can easily be adapted to other national settings, for LMIC or even HICs. Similar models could also be applied in other contexts besides the medical device sector, after further testing, like for instance, in agriculture.
In this study, a group of 10 diseases (top five communicable and top five non-communicable diseases) was selected. For other broader studies, this number can also be higher, thus resulting in different scale of results. Despite the high incidence of communicable diseases, non-communicable diseases account for more than half of the health crisis and are more life-threatening in nature in India. 13 The comparative weight of the type of diseases (communicable versus non-communicable) could also be taken into account when developing national priorities for research. To the best of the author's knowledge, the only comprehensive study to estimate summary measures of population health for the world, by cause, is the Global Burden of Diseases, Injuries, and Risk Factors (GBD) enterprise, which was updated by WHO for the years 2000 and WHO estimates were subsequently updated for the year 2004. 7 Later, WHO developed a comprehensive and consistent set of DALY estimates for years 2000-2012 for population, births, all-cause deaths and specific causes of death as well as WHO estimates for some specific diseases and analyses carried out for the Global Burden of Disease 2010 study. Thus, using data on causes of premature death, loss of health and disability in different populations' mortality and disability, other than a disease, would be valuable to enrich the model itself and the resulting list of priority medical devices. Thus, this methodology could be further improved depending upon the contexts and breadth of the application intended.

CONCLUSION
The findings of this study outline extrapolative projections for the future for population health, 14 based on certain demographic and trade assumptions. We use the word assumption with measured responsibility because irrespective of the plausible uncertainty, this is only an application of epidemiological data and its convergence with macro-economic indicators. This is expected to impact a generation of indigenous manufacturing and innovation that are need-driven, market-driven, as well as highly relevant for self-reliance in the context of pandemics. In turn, such innovation policy could guide the government to make strategic resource allocation, positively impact healthcare indicators besides improving manufacturing and employment. Similar models for other sectors/programs that require to be fueled by innovations is suggested for further research.
Governance of innovation has been influenced by very few methods and decision making is reactive to the understanding of technology at the point of product submission. Such proactive methods, one described through this study, could initiate a wider dialogue on resilient innovation policy which has become so much more pressing in times where all nations simultaneously experience enormous dependencies on the import of crucial medical devices.